Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1435 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Logical Reasoning | KK 6–8 ppl. | 21 | Jun 2, 2026 | ||
| Logical Reasoning | KK 4–5 ppl. | 21 | Jun 2, 2026 | ||
| Logical Reasoning | KK 2–3 ppl. | 21 | Jun 2, 2026 | ||
| Hallucination Detection | SQuAD 2 (test) | 21 | Jun 2, 2026 | ||
| Refusal | MuSiQue Un | 21 | Jun 2, 2026 | ||
| Faithfulness | ConFiQA | 21 | Jun 2, 2026 | ||
| Multi-hop Reasoning | TAT-QA | 21 | Jun 2, 2026 | ||
| Common Sense Reasoning | BBH Sports Understanding | 21 | Jun 2, 2026 | ||
| Mathematical Reasoning | HMMT25 | 21 | Jun 2, 2026 | ||
| Mathematical Reasoning | HMMT25 | 21 | Jun 2, 2026 | ||
| Mathematical Reasoning | AIME 26 | 21 | Jun 2, 2026 | ||
| Long-video understanding | OVOBench Real-Time | 21 | Jun 1, 2026 | ||
| Binary Classification | Wisconsin Breast Cancer | 21 | Jul 7, 2026 | ||
| Long Video Understanding | VideoMME Overall (w/o sub.) 2025 | 21 | Jun 1, 2026 | ||
| Remaining Useful Life Estimation | C-MAPSS FD004 (test) | 21 | Jun 1, 2026 | ||
| Remaining Useful Life Estimation | C-MAPSS FD003 (test) | 21 | Jun 1, 2026 | ||
| Single-Hop Search-augmented Question Answering | PopQA | 21 | Jun 11, 2026 | ||
| Single-Hop Search-augmented Question Answering | TriviaQA | 21 | Jun 11, 2026 | ||
| Single-Hop Search-augmented Question Answering | NQ | 21 | Jun 11, 2026 | ||
| Metric Depth Estimation | ETH3D | 21 | Jun 1, 2026 | ||
| Question Answering | SQuAD | 21 | Jun 1, 2026 | ||
| Language Understanding | MMLU | 21 | Jun 1, 2026 | ||
| Spatial Reasoning | All-Angles Bench | 21 | May 29, 2026 | ||
| Code Execution | Live Code Execution | 21 | May 29, 2026 | ||
| Code Generation | Live Code Generation | 21 | May 29, 2026 |