Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 512 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Agentic Oversight | Agent-SafetyBench | 42 | Feb 26, 2026 | ||
| Retrieval-Augmented Generation | LOFT | 42 | Feb 26, 2026 | ||
| General Language Evaluation | Aggregated MMLU, BoolQ, OpenBookQA, RTE | 42 | Apr 9, 2026 | ||
| Node classification | CHAMELEON FILTERED (test) | 42 | Jun 4, 2026 | ||
| Node classification | SQUIRREL FILTERED (test) | 42 | Jun 4, 2026 | ||
| Quantum Property Prediction | QM9 | 42 | May 12, 2026 | ||
| Machine Translation | FLORES (test) | 42 | Jun 30, 2026 | ||
| Helpfulness | Alpaca Eval | 42 | Jul 7, 2026 | ||
| Model Helpfulness Evaluation | Just-Eval (test) | 42 | Feb 26, 2026 | ||
| Question Answering | HotpotQA | 42 | Mar 18, 2026 | ||
| AI-generated text detection | RAID (test) | 42 | Feb 26, 2026 | ||
| Citation-augmented Question Answering | bar-GT, PK 1.0 (test) | 42 | Feb 27, 2026 | ||
| Geometric Reasoning | Geometry3K | 42 | Apr 3, 2026 | ||
| Image Editing | AnyEdit (test) | 42 | Jun 16, 2026 | ||
| Video Question Answering | LongVideoBench (test) | 42 | Apr 23, 2026 | ||
| 6D Object Pose Estimation | BOP Core Datasets Challenge (test) | 42 | Feb 26, 2026 | ||
| Multimodal Understanding and Reasoning | MMMU, SEED, OCRBench, VizWiz, ScienceQA, TextVQA (test/val) | 42 | Feb 26, 2026 | ||
| Infrared and visible image fusion | RoadScene | 42 | May 14, 2026 | ||
| Pansharpening | GF2 full-resolution (test) | 42 | May 19, 2026 | ||
| Image Classification | 11-Dataset Average | 42 | Mar 17, 2026 | ||
| Image Super-Resolution | Set14 x4 scale (test) | 42 | Jun 19, 2026 | ||
| Natural Language Processing | T0 MTest11 P3 (test) | 42 | Feb 26, 2026 | ||
| Complex Reasoning | BBH (val) | 42 | Feb 26, 2026 | ||
| Code Generation | LiveCodeBench (test) | 42 | May 12, 2026 | ||
| Compositional Image Retrieval | FashionIQ 1.0 (val) | 42 | Feb 26, 2026 |