{"task":"Code Generation","dataset":"Terminal-Bench Hard","metric_names":["Accuracy"],"rows":[{"id":759610,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Sol (Max) (OpenAI)","metrics":{"Accuracy":"65.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759552,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Fable 5 (Max, Opus 4.8 Fallback) (Anthropic)","metrics":{"Accuracy":"62.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759932,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Sol (Medium) (OpenAI)","metrics":{"Accuracy":"62.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760004,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Terra (Xhigh) (OpenAI)","metrics":{"Accuracy":"62.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759764,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Sol (High) (OpenAI)","metrics":{"Accuracy":"62.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759719,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Sol (Xhigh) (OpenAI)","metrics":{"Accuracy":"61.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759978,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 (Xhigh) (OpenAI)","metrics":{"Accuracy":"60.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760214,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Sol (Low) (OpenAI)","metrics":{"Accuracy":"60.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760035,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 (High) (OpenAI)","metrics":{"Accuracy":"59.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759812,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.8 (Max) (Anthropic)","metrics":{"Accuracy":"58.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759791,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Terra (Max) (OpenAI)","metrics":{"Accuracy":"57.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759952,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 (Xhigh) (OpenAI)","metrics":{"Accuracy":"57.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760126,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Terra (High) (OpenAI)","metrics":{"Accuracy":"57.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760156,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 (Medium) (OpenAI)","metrics":{"Accuracy":"57.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760296,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.7 (Non-reasoning, High) (Anthropic)","metrics":{"Accuracy":"54.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760396,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3.1 Pro Preview (Google)","metrics":{"Accuracy":"53.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760251,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.3 Codex (Xhigh) (OpenAI)","metrics":{"Accuracy":"53"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760379,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Sonnet 4.6 (Max) (Anthropic)","metrics":{"Accuracy":"53"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760304,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 (Low) (OpenAI)","metrics":{"Accuracy":"52.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760920,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 mini (Xhigh) (OpenAI)","metrics":{"Accuracy":"52.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":759847,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.7 (Max) (Anthropic)","metrics":{"Accuracy":"51.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760177,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5.2 (Max) (Z AI)","metrics":{"Accuracy":"50.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760408,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.7 Max (Alibaba)","metrics":{"Accuracy":"50.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761017,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"49.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761228,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"KAT Coder Pro V2 (KwaiKAT)","metrics":{"Accuracy":"49.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760632,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.6 (Non-reasoning, High) (Anthropic)","metrics":{"Accuracy":"48.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760338,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.2 (Xhigh) (OpenAI)","metrics":{"Accuracy":"47"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760439,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.5 (Reasoning) (Anthropic)","metrics":{"Accuracy":"47"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760761,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.7 Plus (Alibaba)","metrics":{"Accuracy":"47"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760274,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.6 (Max) (Anthropic)","metrics":{"Accuracy":"46.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760324,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Pro 0424 (Max) (DeepSeek)","metrics":{"Accuracy":"46.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760852,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Sonnet 4.6 (Non-reasoning, High) (Anthropic)","metrics":{"Accuracy":"46.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760943,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3.5 Flash (Minimal) (Google)","metrics":{"Accuracy":"46.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760287,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Muse Spark (Meta)","metrics":{"Accuracy":"45.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760840,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.1 (High) (OpenAI)","metrics":{"Accuracy":"45.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760696,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2.7 Code (Kimi)","metrics":{"Accuracy":"44.7"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760468,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 Max Preview (Alibaba)","metrics":{"Accuracy":"43.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760547,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.6 Terra (Low) (OpenAI)","metrics":{"Accuracy":"43.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760572,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 Plus (Alibaba)","metrics":{"Accuracy":"43.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760584,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2.6 (Reasoning) (Kimi)","metrics":{"Accuracy":"43.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760521,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5 (Reasoning) (Z AI)","metrics":{"Accuracy":"43.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760529,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 (Low) (OpenAI)","metrics":{"Accuracy":"43.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760614,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.2 (Medium) (OpenAI)","metrics":{"Accuracy":"43.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760663,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5.1 (Reasoning) (Z AI)","metrics":{"Accuracy":"43.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760684,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2.5-Pro (Reasoning) (Xiaomi)","metrics":{"Accuracy":"43.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760425,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax-M3 (MiniMax)","metrics":{"Accuracy":"42.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761001,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Sonnet 4.6 (Non-reasoning, Low) (Anthropic)","metrics":{"Accuracy":"42.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761117,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.5 Instant (May 2026) (OpenAI)","metrics":{"Accuracy":"42.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761356,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 nano (Xhigh) (OpenAI)","metrics":{"Accuracy":"42.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760347,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Pro 0424 (High) (DeepSeek)","metrics":{"Accuracy":"41.7"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760512,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3 Pro Preview (High) (Google)","metrics":{"Accuracy":"41.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760749,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2.5 (Xiaomi)","metrics":{"Accuracy":"41.7"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760240,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3.5 Flash (High) (Google)","metrics":{"Accuracy":"40.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760448,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Pro (Xiaomi)","metrics":{"Accuracy":"40.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760740,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.20 0309 (Reasoning) (SpaceXAI)","metrics":{"Accuracy":"40.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760966,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude Opus 4.5 (Non-reasoning) (Anthropic)","metrics":{"Accuracy":"40.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761651,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 397B A17B (Reasoning) (Alibaba)","metrics":{"Accuracy":"40.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760197,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3.5 Flash (Medium) (Google)","metrics":{"Accuracy":"39.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761098,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax-M2.7 (MiniMax)","metrics":{"Accuracy":"39.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761220,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5 (Non-reasoning) (Z AI)","metrics":{"Accuracy":"39.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760643,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3 Flash Preview (Reasoning) (Google)","metrics":{"Accuracy":"38.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760875,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Flash 0420 (High) (DeepSeek)","metrics":{"Accuracy":"38.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760716,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.20 0309 v2 (Reasoning) (SpaceXAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760800,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 Codex (High) (OpenAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760810,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.3 (High) (SpaceXAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760975,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2.6 (Non-reasoning) (Kimi)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761068,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 (Medium) (OpenAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761137,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4 (SpaceXAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761682,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"37.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760460,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.2 Codex (Xhigh) (OpenAI)","metrics":{"Accuracy":"37.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761456,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"o3 (OpenAI)","metrics":{"Accuracy":"37.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761045,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nemotron 3 Ultra 550B A55B (Reasoning) (NVIDIA)","metrics":{"Accuracy":"36.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-04","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761314,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Pro 0424 (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"36.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762161,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 31B (Reasoning) (Google)","metrics":{"Accuracy":"36.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760772,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Omni-0327 (Xiaomi)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760899,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5.1 (Non-reasoning) (Z AI)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760908,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Flash 0420 (Max) (DeepSeek)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761247,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.2 (Reasoning) (DeepSeek)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761266,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 397B A17B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761371,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4.5 Sonnet (Reasoning) (Anthropic)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761519,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Step 3.7 Flash (StepFun)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761662,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2.5-Pro (Non-reasoning) (Xiaomi)","metrics":{"Accuracy":"35.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760936,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Omni (Xiaomi)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760954,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.1 Codex (High) (OpenAI)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760990,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2.5 (Reasoning) (Kimi)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761089,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax-M2.5 (MiniMax)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761275,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 27B (Reasoning) (Alibaba)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761671,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 35B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762203,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.2 Speciale (DeepSeek)","metrics":{"Accuracy":"34.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761080,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4.1 Opus (Reasoning) (Anthropic)","metrics":{"Accuracy":"34.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761109,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hy3-preview (Reasoning) (Tencent)","metrics":{"Accuracy":"34.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761164,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3 Pro Preview (Low) (Google)","metrics":{"Accuracy":"34.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761502,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 mini (Medium) (OpenAI)","metrics":{"Accuracy":"34.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761590,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V4 Flash 0420 (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"34.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760604,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-5-Turbo (Z AI)","metrics":{"Accuracy":"33.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761427,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.1 Codex mini (High) (OpenAI)","metrics":{"Accuracy":"33.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761483,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 nano (Medium) (OpenAI)","metrics":{"Accuracy":"33.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761826,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 mini (High) (OpenAI)","metrics":{"Accuracy":"33.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762243,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Medium 3.5 (Mistral)","metrics":{"Accuracy":"33.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760983,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM 5V Turbo (Reasoning) (Z AI)","metrics":{"Accuracy":"32.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761032,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 (High) (OpenAI)","metrics":{"Accuracy":"32.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761056,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 27B (Reasoning) (Alibaba)","metrics":{"Accuracy":"32.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761787,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Step 3.5 Flash 2603 (StepFun)","metrics":{"Accuracy":"32.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761962,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.2 (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"32.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761176,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.7 (Reasoning) (Z AI)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761538,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 27B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761705,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3 Flash Preview (Non-reasoning) (Google)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761758,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hy3-preview (Non-reasoning) (Tencent)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761777,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.2 (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762304,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.1 Terminus (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"31.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761146,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Flash (Feb 2026) (Xiaomi)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761200,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2 Thinking (Kimi)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761387,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4 Opus (Reasoning) (Anthropic)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761579,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4 Sonnet (Reasoning) (Anthropic)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761766,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ling-2.6-1T (InclusionAI)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761908,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.2 Exp (Reasoning) (DeepSeek)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761992,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 122B A10B (Reasoning) (Alibaba)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763284,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"North Mini Code (Cohere)","metrics":{"Accuracy":"31.1"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760821,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.3 (Medium) (SpaceXAI)","metrics":{"Accuracy":"30.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761749,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.7 (Non-reasoning) (Z AI)","metrics":{"Accuracy":"30.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762148,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.1 Terminus (Reasoning) (DeepSeek)","metrics":{"Accuracy":"30.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762313,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 31B (Non-reasoning) (Google)","metrics":{"Accuracy":"30.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761714,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 122B A10B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"29.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761305,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax-M2.1 (MiniMax)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761399,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 mini (Medium) (OpenAI)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761549,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4.5 Sonnet (Non-reasoning) (Anthropic)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761598,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"JT-35B-Flash (China Mobile)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761864,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ring-2.6-1T (InclusionAI)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762118,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.6 (Non-reasoning) (Z AI)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762540,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nemotron 3 Super 120B A12B (Reasoning) (NVIDIA)","metrics":{"Accuracy":"28.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761325,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Flash (Reasoning) (Xiaomi)","metrics":{"Accuracy":"28"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761810,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4.5 Haiku (Reasoning) (Anthropic)","metrics":{"Accuracy":"27.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761875,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Step 3.5 Flash (StepFun)","metrics":{"Accuracy":"27.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761897,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4 Sonnet (Non-reasoning) (Anthropic)","metrics":{"Accuracy":"27.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762038,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 4.5 Haiku (Non-reasoning) (Anthropic)","metrics":{"Accuracy":"27.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":760886,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.3 (Low) (SpaceXAI)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761338,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 (Low) (OpenAI)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761558,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 35B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761798,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Doubao Seed Code (ByteDance Seed)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761934,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Pro (Google)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762348,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mercury 2 (Inception)","metrics":{"Accuracy":"26.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761608,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax-M2 (MiniMax)","metrics":{"Accuracy":"25.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761949,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiMo-V2-Flash (Non-reasoning) (Xiaomi)","metrics":{"Accuracy":"25.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762072,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 35B A3B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"25.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761633,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.6 (Reasoning) (Z AI)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762227,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"ERNIE 5.0 Thinking Preview (Baidu)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762327,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.2 Exp (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762386,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.1 (Reasoning) (DeepSeek)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762463,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Command A+ (Cohere)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762473,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 26B A4B (Non-reasoning) (Google)","metrics":{"Accuracy":"25"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761286,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Max Thinking (Alibaba)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761439,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.1 Fast (Reasoning) (SpaceXAI)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762004,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 3.1 Flash-Lite (Google)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762267,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Pro Preview (Medium) (Amazon)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762362,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3.1 (Non-reasoning) (DeepSeek)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762822,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 nano (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762987,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 9B (Reasoning) (Alibaba)","metrics":{"Accuracy":"24.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762050,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2 0905 (Kimi)","metrics":{"Accuracy":"23.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762799,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"HyperNova 60B 2605 (High, Based on gpt-oss-120b) (Multiverse Computing)","metrics":{"Accuracy":"23.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762850,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"gpt-oss-120b (High) (OpenAI)","metrics":{"Accuracy":"23.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762214,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K-EXAONE (Reasoning) (LG AI Research)","metrics":{"Accuracy":"22.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-31","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762434,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.1 (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"22.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763104,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Trinity Large Thinking (Arcee AI)","metrics":{"Accuracy":"22.7"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-01","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762127,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.7-Flash (Reasoning) (Z AI)","metrics":{"Accuracy":"22"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762184,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.20 0309 (Non-reasoning) (SpaceXAI)","metrics":{"Accuracy":"22"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762569,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.5 (Reasoning) (Z AI)","metrics":{"Accuracy":"22"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761416,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 Omni Plus (Alibaba)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761491,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.6 27B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761729,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 3.7 Sonnet (Reasoning) (Anthropic)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762063,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 3.7 Sonnet (Non-reasoning) (Anthropic)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762830,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nemotron Cascade 2 30B A3B (NVIDIA)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763371,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ling 2.6 Flash (InclusionAI)","metrics":{"Accuracy":"21.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761982,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Max (Alibaba)","metrics":{"Accuracy":"20.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762453,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"EXAONE 4.5 33B (Reasoning) (LG AI Research)","metrics":{"Accuracy":"20.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763021,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.5-Air (Z AI)","metrics":{"Accuracy":"20.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762623,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Max (Preview) (Alibaba)","metrics":{"Accuracy":"19.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761530,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2.5 (Non-reasoning) (Kimi)","metrics":{"Accuracy":"18.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761693,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4 Fast (Reasoning) (SpaceXAI)","metrics":{"Accuracy":"18.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762290,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.3 (Non-reasoning) (SpaceXAI)","metrics":{"Accuracy":"18.9"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762749,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Coder 480B A35B Instruct (Alibaba)","metrics":{"Accuracy":"18.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763826,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Devstral 2 (Mistral)","metrics":{"Accuracy":"18.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762254,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 12B (Reasoning) (Google)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762443,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 9B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762481,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 4B (Reasoning) (Alibaba)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762700,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"JT-MINI (China Mobile)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762882,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 (Minimal) (OpenAI)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762998,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5.4 mini (Non-reasoning) (OpenAI)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763501,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Coder Next (Alibaba)","metrics":{"Accuracy":"18.2"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761920,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Max Thinking (Preview) (Alibaba)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762175,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 3 mini Reasoning (High) (SpaceXAI)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-02-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762281,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok Code Fast 1 (SpaceXAI)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762554,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Pro Preview (Low) (Amazon)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762635,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Lite (Medium) (Amazon)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762647,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 nano (Medium) (OpenAI)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762938,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Small 4 (Reasoning) (Mistral)","metrics":{"Accuracy":"17.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762027,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash Preview (Sep '25) (Reasoning) (Google)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762234,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.20 0309 v2 (Non-reasoning) (SpaceXAI)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762420,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Lite (High) (Amazon)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763248,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Pro Preview (Non-reasoning) (Amazon)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764358,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Devstral Small 2 (Mistral)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766149,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Cogito v2.1 (Reasoning) (Deep Cogito)","metrics":{"Accuracy":"16.7"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762496,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek R1 0528 (May '25) (DeepSeek)","metrics":{"Accuracy":"15.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762581,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi K2 (Kimi)","metrics":{"Accuracy":"15.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763488,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Large 3 (Mistral)","metrics":{"Accuracy":"15.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761854,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"o4-mini (High) (OpenAI)","metrics":{"Accuracy":"15.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762736,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 235B A22B 2507 Instruct (Alibaba)","metrics":{"Accuracy":"15.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763359,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3 0324 (DeepSeek)","metrics":{"Accuracy":"15.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763404,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Coder 30B A3B Instruct (Alibaba)","metrics":{"Accuracy":"15.2"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-31","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762410,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Apriel-v1.6-15B-Thinker (ServiceNow)","metrics":{"Accuracy":"14.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762687,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning) (Google)","metrics":{"Accuracy":"14.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762929,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4.1 Fast (Non-reasoning) (SpaceXAI)","metrics":{"Accuracy":"14.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762952,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.6V (Reasoning) (Z AI)","metrics":{"Accuracy":"14.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763297,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 mini (Minimal) (OpenAI)","metrics":{"Accuracy":"14.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761839,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 26B A4B (Reasoning) (Google)","metrics":{"Accuracy":"13.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762509,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash (Reasoning) (Google)","metrics":{"Accuracy":"13.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762595,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 235B A22B 2507 (Reasoning) (Alibaba)","metrics":{"Accuracy":"13.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762611,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-4.1 (OpenAI)","metrics":{"Accuracy":"13.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763726,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) (NVIDIA)","metrics":{"Accuracy":"13.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762087,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"o1 (OpenAI)","metrics":{"Accuracy":"12.9"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762773,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Magistral Medium 1.2 (Mistral)","metrics":{"Accuracy":"12.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763148,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 (ChatGPT) (OpenAI)","metrics":{"Accuracy":"12.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763167,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning) (Google)","metrics":{"Accuracy":"12.9"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762521,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 nano (High) (OpenAI)","metrics":{"Accuracy":"12.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762918,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"HyperCLOVA X SEED Think (32B) (Naver)","metrics":{"Accuracy":"12.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763042,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 4 Fast (Non-reasoning) (SpaceXAI)","metrics":{"Accuracy":"12.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763321,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash (Non-reasoning) (Google)","metrics":{"Accuracy":"12.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764837,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning) (NVIDIA)","metrics":{"Accuracy":"12.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762398,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 235B A22B (Reasoning) (Alibaba)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762711,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Grok 3 (SpaceXAI)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-02-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763115,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 4B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763464,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 12B (Non-reasoning) (Google)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2026-06-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764408,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hermes 4 - Llama-3.1 405B (Reasoning) (Nous Research)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764551,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Kimi Linear 48B A3B Instruct (Kimi)","metrics":{"Accuracy":"11.4"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762096,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 35B A3B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-24","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762339,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Apriel-v1.5-15B-Thinker (ServiceNow)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762871,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LongCat Flash Lite (LongCat)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763527,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Medium 3.1 (Mistral)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763541,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ling-1T (InclusionAI)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763636,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Small 4 (Non-reasoning) (Mistral)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763655,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"gpt-oss-20b (High) (OpenAI)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764210,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ling-flash-2.0 (InclusionAI)","metrics":{"Accuracy":"10.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762974,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Next 80B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"9.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763309,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K2-V2 (High) (Institute of Foundation Models)","metrics":{"Accuracy":"9.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764472,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hermes 4 - Llama-3.1 405B (Non-reasoning) (Nous Research)","metrics":{"Accuracy":"9.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":761620,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"KAT-Coder-Pro V1 (KwaiKAT)","metrics":{"Accuracy":"9.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763129,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"INTELLECT-3 (Based on GLM-4.5-Air) (Prime Intellect)","metrics":{"Accuracy":"9.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763581,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Magistral Medium 1 (Mistral)","metrics":{"Accuracy":"9.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763628,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Devstral Medium (Mistral)","metrics":{"Accuracy":"9.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762656,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 Omni Flash (Alibaba)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763176,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nemotron 3 Nano Omni 30B A3B Reasoning (NVIDIA)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763605,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K2-V2 (Medium) (Institute of Foundation Models)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763713,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 E4B (Reasoning) (Google)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763903,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-4o (Nov '24) (OpenAI)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2024-11-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763932,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 32B Instruct (Alibaba)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764243,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-4o (Aug '24) (OpenAI)","metrics":{"Accuracy":"8.3"},"paper_url":null,"paper_title":null,"paper_date":"2024-08-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762761,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 32B (Reasoning) (Alibaba)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763205,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-4.1 mini (OpenAI)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763384,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Next 80B A3B Instruct (Alibaba)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763475,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning) (Google)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764196,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Solar Pro 3 (Upstage)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764440,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 E4B (Non-reasoning) (Google)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764654,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ring-flash-2.0 (InclusionAI)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764689,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Small 3.1 (Mistral)","metrics":{"Accuracy":"7.6"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762667,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"o3-mini (OpenAI)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-01-31","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762723,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Seed-OSS-36B-Instruct (ByteDance Seed)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762862,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K2 Think V2 (Institute of Foundation Models)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762963,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K-EXAONE (Non-reasoning) (LG AI Research)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-31","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763089,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ring-1T (InclusionAI)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763222,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 4 Maverick (Meta)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763274,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 235B A22B Instruct (Alibaba)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763553,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova Premier (Amazon)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763775,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Lite (Non-reasoning) (Amazon)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763887,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek V3 (Dec '24) (DeepSeek)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763980,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Small 3.2 (Mistral)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764054,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Omni (Non-reasoning) (Amazon)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764563,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.1 Instruct 405B (Meta)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-07-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764711,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-5 nano (Minimal) (OpenAI)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764910,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.5V (Non-reasoning) (Z AI)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764985,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 30B A3B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766113,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron 3 Nano 4B (NVIDIA)","metrics":{"Accuracy":"6.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762895,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek R1 (Jan '25) (DeepSeek)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-01-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763075,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"o3-mini (High) (OpenAI)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-01-31","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763425,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 235B A22B (Reasoning) (Alibaba)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763788,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Devstral Small (May '25) (Mistral)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-21","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763966,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 235B A22B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764102,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 30B A3B Instruct (Alibaba)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764316,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Devstral Small (Jul '25) (Mistral)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764347,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Large 2 (Nov '24) (Mistral)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2024-11-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764384,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 30B A3B 2507 Instruct (Alibaba)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764397,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"ERNIE 4.5 300B A47B (Baidu)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764762,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova Pro (Amazon)","metrics":{"Accuracy":"6.1"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763189,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"gpt-oss-120b (Low) (OpenAI)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763335,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 30B A3B 2507 (Reasoning) (Alibaba)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-30","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763448,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 30B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763617,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama Nemotron Super 49B v1.5 (Reasoning) (NVIDIA)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764270,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Step3 VL 10B (StepFun)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764334,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.5V (Reasoning) (Z AI)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764891,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 14B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"5.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762374,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Omni (Medium) (Amazon)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763262,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"gpt-oss-20b (Low) (OpenAI)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763840,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Magistral Small 1.2 (Mistral)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763873,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash-Lite (Reasoning) (Google)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763996,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Magistral Small 1 (Mistral)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764113,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hermes 4 - Llama-3.1 70B (Reasoning) (Nous Research)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764255,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen2.5 Instruct 72B (Alibaba)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2024-09-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764432,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron Nano 12B v2 VL (Reasoning) (NVIDIA)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764529,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"K2-V2 (Low) (Institute of Foundation Models)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764622,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2.5-8B-A1B (Liquid AI)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764750,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Solar Pro 2 (Non-reasoning) (Upstage)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764797,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.1 Nemotron Instruct 70B (NVIDIA)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2024-10-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764872,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 4B 2507 Instruct (Alibaba)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765211,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ministral 3 14B (Mistral)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765575,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ministral 3 8B (Mistral)","metrics":{"Accuracy":"4.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762790,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Lite (Low) (Amazon)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763008,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova 2.0 Omni (Low) (Amazon)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763137,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.7-Flash (Non-reasoning) (Z AI)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-19","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763515,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Motif-2-12.7B-Reasoning (Motif Technologies)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-04","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763593,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mistral Medium 3 (Mistral)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763679,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.0 Flash (Feb '25) (Google)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-02-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764015,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"EXAONE 4.0 32B (Reasoning) (LG AI Research)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764025,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 8B (Reasoning) (Alibaba)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764039,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 14B (Reasoning) (Alibaba)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764183,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GPT-4.1 nano (OpenAI)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764231,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Omni 30B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764492,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama Nemotron Super 49B v1.5 (Non-reasoning) (NVIDIA)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764784,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 2B (Reasoning) (Alibaba)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765152,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 2B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765302,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Phi-4 (Microsoft)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765930,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3 27B Instruct (Google)","metrics":{"Accuracy":"3.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":762813,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax M1 80k (MiniMax)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763857,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 32B (Reasoning) (Alibaba)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763953,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"GLM-4.6V (Non-reasoning) (Z AI)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764219,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 E2B (Reasoning) (Google)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764281,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.3 Instruct 70B (Meta)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764421,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Solar Pro 2 (Reasoning) (Upstage)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765009,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.1 Instruct 70B (Meta)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2024-07-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766160,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mi:dm K 2.5 Pro Preview (Korea Telecom)","metrics":{"Accuracy":"3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763062,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Mi:dm K 2.5 Pro (Korea Telecom)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763156,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Solar Open 100B (Reasoning) (Upstage)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763238,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniMax M1 40k (MiniMax)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763393,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Tri-21B-think Preview (Trillion Labs)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763691,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 3.5 Haiku (Anthropic)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2024-10-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764171,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Falcon-H1R-7B (TII UAE)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-04","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764296,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 30B A3B (Reasoning) (Alibaba)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764373,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.1 Nemotron Ultra 253B v1 (Reasoning) (NVIDIA)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764448,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.1 30B (IBM)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764576,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 8B (Reasoning) (Alibaba)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764602,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 8B Instruct (Alibaba)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764935,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemini 2.5 Flash-Lite (Non-reasoning) (Google)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765026,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Sarvam 30B (High) (Sarvam)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765059,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 4 E2B (Non-reasoning) (Google)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765187,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Jamba 1.7 Large (AI21 Labs)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765198,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 H Small (IBM)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765255,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 8B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765331,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.1 3B (IBM)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765648,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Sarvam M (Reasoning, Based on Mistral Small 3.1) (Sarvam)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766008,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3n E4B Instruct (Google)","metrics":{"Accuracy":"2.3"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763741,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 4B 2507 (Reasoning) (Alibaba)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763756,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Sarvam 105B (High) (Sarvam)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-06","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764067,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek R1 0528 Qwen3 8B (DeepSeek)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-05-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764080,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 4 Scout (Meta)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764140,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"DeepSeek R1 Distill Llama 70B (DeepSeek)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-01-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764460,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron Nano 9B V2 (Reasoning) (NVIDIA)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764730,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 4B (Reasoning) (Alibaba)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765069,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Olmo 3 32B Think (Allen Institute for AI)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765145,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"EXAONE 4.0 32B (Non-reasoning) (LG AI Research)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765234,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 Omni 30B A3B Instruct (Alibaba)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765322,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova Micro (Amazon)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765830,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 Micro (IBM)","metrics":{"Accuracy":"1.5"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-22","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763644,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Tri-21B-Think (Trillion Labs)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764775,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Command A (Cohere)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764809,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.1 Instruct 8B (Meta)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-07-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764849,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron Nano 9B V2 (Non-reasoning) (NVIDIA)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764948,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nova Lite (Amazon)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-12-03","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765439,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Jamba Reasoning 3B (AI21 Labs)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-08","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765476,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Olmo 3 7B Think (Allen Institute for AI)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765495,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Claude 3 Haiku (Anthropic)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-03-04","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765527,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ling-mini-2.0 (InclusionAI)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-09","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765591,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3 Instruct 70B (Meta)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-04-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765612,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.2 Instruct 11B (Vision) (Meta)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2024-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765747,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2 2.6B (Liquid AI)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765981,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3 4B Instruct (Google)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765996,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3n E2B Instruct (Google)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-06-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766127,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3 12B Instruct (Google)","metrics":{"Accuracy":"0.8"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763704,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.3 Nemotron Super 49B v1 (Reasoning) (NVIDIA)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763749,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniCPM5-1B (Reasoning) (OpenBMB)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763804,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniCPM5-1B (Non-reasoning) (OpenBMB)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":763912,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Nanbeige4.1-3B (Nanbeige)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764592,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.3 Nemotron Super 49B v1 (Non-reasoning) (NVIDIA)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764676,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Olmo 3.1 32B Think (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-12","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":764973,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Hermes 4 - Llama-3.1 70B (Non-reasoning) (Nous Research)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-27","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765017,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.1 8B (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-04-29","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765037,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Olmo 3.1 32B Instruct (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765131,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Phi-4 Mini Instruct (Microsoft)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-02-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765168,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 0.8B (Reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765269,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"OLMo 2 32B (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765284,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2 24B A2B (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765353,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Phi-3 Mini Instruct 3.8B (Microsoft)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-04-23","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765363,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) (NVIDIA)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765427,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"MiniCPM-V 4.6 1.3B (OpenBMB)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-05-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765450,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 VL 4B Instruct (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765464,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Reka Flash 3 (Reka AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765506,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"OLMo 2 7B (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-11-26","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765517,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Molmo 7B-D (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765621,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3.5 0.8B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-03-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765694,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Exaone 4.0 1.2B (Reasoning) (LG AI Research)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765705,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Olmo 3 7B Instruct (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-11-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765717,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Exaone 4.0 1.2B (Non-reasoning) (LG AI Research)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-15","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765725,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2.5-1.2B-Thinking (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-20","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765736,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Jamba 1.7 Mini (AI21 Labs)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765755,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2.5-1.2B-Instruct (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765773,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 H 1B (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765785,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 1.7B (Reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765811,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3 270M (Google)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-08-14","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765819,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Apertus 70B Instruct (Swiss AI Initiative)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765876,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 1B (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765884,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Molmo2-8B (Allen Institute for AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-11","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765894,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2 8B A1B (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-07","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765905,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 3.3 8B (Non-reasoning) (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-16","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765917,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 1.7B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765945,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Ministral 3 3B (Mistral)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-12-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765957,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Apertus 8B Instruct (Swiss AI Initiative)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-09-02","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":765969,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Gemma 3 1B Instruct (Google)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-03-13","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766022,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 350M (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766032,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Granite 4.0 H 350M (IBM)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-10-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766043,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2 1.2B (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-07-10","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766051,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"LFM2.5-VL-1.6B (Liquid AI)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-01-05","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766062,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3 Instruct 8B (Meta)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-04-18","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766074,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Llama 3.2 Instruct 1B (Meta)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2024-09-25","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766085,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 0.6B (Non-reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766097,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Qwen3 0.6B (Reasoning) (Alibaba)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2025-04-28","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]},{"id":766105,"task":"Code Generation","parent_task":null,"dataset":"Terminal-Bench Hard","model_name":"Tiny Aya Global (Cohere)","metrics":{"Accuracy":"0"},"paper_url":null,"paper_title":null,"paper_date":"2026-02-17","code_links":[{"title":"artificialanalysis.ai","url":"https://artificialanalysis.ai/"}],"metrics_order":"[\"Accuracy\"]","area":"Natural Language Processing","uses_additional_data":null,"source":"external","tags":[]}]}