Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 527 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Fine-Grained Perception | MMVP | 41 | Jun 9, 2026 | ||
| Depth Super-Resolution / Completion | ETH-3D (test) | 41 | Feb 26, 2026 | ||
| Vision-Language Understanding and Reasoning | LLaVA Multimodal Evaluation Suite (GQA, MMBench, MME, POPE, ScienceQA, VQAv2, TextVQA, SEED-Bench, MM-Vet, VizWiz) 1.5 (test/val) | 41 | May 14, 2026 | ||
| Anomaly Detection | BTAD | 41 | Apr 14, 2026 | ||
| Geometry problem solving | Geo3K | 41 | May 21, 2026 | ||
| Text-to-Video Generation | VBench (test) | 41 | Jul 2, 2026 | ||
| Depth Super-Resolution | NYU v2 | 41 | Mar 9, 2026 | ||
| Multi-task Language Understanding | MMLU Pro | 41 | May 19, 2026 | ||
| Agent | τ2-Bench | 41 | May 26, 2026 | ||
| Large Language Model Evaluation | Open LLM Leaderboard | 41 | Apr 14, 2026 | ||
| Question Answering | MathQA (test) | 41 | Mar 12, 2026 | ||
| Code Generation | BigCodeBench-Completion Full | 41 | Feb 26, 2026 | ||
| Multi-turn conversation | MT-Bench | 41 | Mar 10, 2026 | ||
| Document Retrieval | MIRACL (dev) | 41 | Feb 26, 2026 | ||
| Document Retrieval | MSMarco (dev) | 41 | Feb 26, 2026 | ||
| Instruction Following | Multi-IF | 41 | Jun 9, 2026 | ||
| Code Generation | HumanEval | 41 | Feb 26, 2026 | ||
| Mathematical Reasoning | MATH500 | 41 | Feb 26, 2026 | ||
| Science Simulation | SciWorld | 41 | Feb 26, 2026 | ||
| Web Navigation Task Success | MIND2WEB ONLINE (test) | 41 | May 15, 2026 | ||
| Composed Image Retrieval | FashionIQ Toptee | 41 | Jul 3, 2026 | ||
| Mathematical Reasoning | AIME25 | 41 | May 11, 2026 | ||
| Agentic Task Success | MCP-Universe | 41 | Jun 16, 2026 | ||
| Audio-Visual Question Answering | MUSIC-AVQA-R (test) | 41 | Jul 7, 2026 | ||
| Passage-pair linking | Literary analysis corpus (Goodreads, Sparknotes, Litcharts) | 41 | Feb 26, 2026 |