@article{odysseybenchevaluatingllmagentsonlonghor,
  title = {OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows},
  author = {Weixuan Wang and Dongge Han and Daniel Madrigal Diaz and Jin Xu and Victor Rühle and Saravan Rajmohan},
  year = {2025},
  eprint = {2508.09124},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2508.09124},
}