Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 709 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Mathematical Reasoning | AMC 23 | 34 | Jun 12, 2026 | ||
| Audiovisual Video Captioning | UGC-VideoCap | 34 | Jun 9, 2026 | ||
| Retrieval | MLDR (test) | 34 | Feb 26, 2026 | ||
| Mathematical Reasoning | MATH500 | 34 | Feb 26, 2026 | ||
| Mathematical Reasoning | MATH500 | 34 | Feb 26, 2026 | ||
| Question Answering | MuSiQue | 34 | Mar 4, 2026 | ||
| Web Browsing | BrowseComp-zh | 34 | Jun 4, 2026 | ||
| Automatic Speech Recognition | Fleurs zh | 34 | May 25, 2026 | ||
| Process-outcome alignment | PRIME Average | 34 | Feb 26, 2026 | ||
| Process-outcome alignment | PRIME Biology | 34 | Feb 26, 2026 | ||
| Process-outcome alignment | PRIME Chemistry | 34 | Feb 26, 2026 | ||
| Process-outcome alignment | PRIME Physics | 34 | Feb 26, 2026 | ||
| Process-outcome alignment | PRIME Math | 34 | Feb 26, 2026 | ||
| Theorem-based Reasoning | TheoremQA | 34 | Feb 26, 2026 | ||
| General Reasoning | GPQA diamond | 34 | Mar 4, 2026 | ||
| Scientific Reasoning | SuperGPQA | 34 | Apr 14, 2026 | ||
| Broad Information Seeking | WideSearch | 34 | May 1, 2026 | ||
| NER | Marathi NER (test) | 34 | Feb 26, 2026 | ||
| Sentiment Classification | Marathi (test) | 34 | Feb 26, 2026 | ||
| Visual Grounding | DIOR-RSVG | 34 | Apr 10, 2026 | ||
| Goal-oriented Dialogue | Restaurant | 34 | Apr 15, 2026 | ||
| Graph-level molecular classification | BACE | 34 | Jun 30, 2026 | ||
| Abstractive Summarization | XLSum | 34 | Apr 6, 2026 | ||
| 3D Talking Head Generation | DualTalk (test) | 34 | Apr 21, 2026 | ||
| Multi-hop Question Answering | 2Wiki (test) | 34 | Jun 16, 2026 |