Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1362 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Fine-grained Video Understanding | EventHallusion | 22 | Apr 24, 2026 | ||
| Multitask Language Understanding | GMMLU c | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | SciQA (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | StrQA (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | Calendar (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | Meetings (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | Trips (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | ProofNet (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | GSM8k (test) | 22 | Apr 23, 2026 | ||
| Step-level correctness assessment | MATH (test) | 22 | Apr 23, 2026 | ||
| Natural Language Understanding | GLUE | 22 | Apr 23, 2026 | ||
| Information Extraction | FDA | 22 | May 6, 2026 | ||
| Mathematical Problem Solving | MATH 500 | 22 | Jun 16, 2026 | ||
| Deep Research | xBench-DS-2505 | 22 | Apr 23, 2026 | ||
| Text-to-Video Generation | StoryEval-Bench 1.0 (test) | 22 | May 12, 2026 | ||
| Vulnerability Detection | PrimeVul Paired (test) | 22 | May 12, 2026 | ||
| Employee Morale Prediction | Firm posts dataset Tmid, 10 posts threshold | 22 | Apr 23, 2026 | ||
| ROA Prediction | Firm posts dataset Tmid 10 posts threshold | 22 | Apr 23, 2026 | ||
| Object Detection | xView | 22 | Apr 23, 2026 | ||
| Optimal Power Flow prediction | DataKit full topology | 22 | Apr 23, 2026 | ||
| Optimal Power Flow prediction | OPFData (full topology) | 22 | Apr 23, 2026 | ||
| Video Question Answering | MF2 | 22 | Apr 21, 2026 | ||
| Video Question Answering | TIMEScope Long | 22 | Apr 21, 2026 | ||
| Video Question Answering | TIMEScope Short | 22 | Apr 21, 2026 | ||
| Orientation estimation | T-LESS (ViVo) | 22 | Apr 21, 2026 |