Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1931 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Mathematical Reasoning | MATH500 | 18 | Feb 26, 2026 | ||
| Mathematical Reasoning | GSM8K | 18 | Feb 26, 2026 | ||
| Competitive Programming Agent Evaluation | ALE Bench | 18 | Jun 26, 2026 | ||
| Mathematical Reasoning | AIME | 18 | Feb 26, 2026 | ||
| Web Browsing and Navigation | WebWalkerQA | 18 | Feb 26, 2026 | ||
| Scientific and General Reasoning | Theorem QA | 18 | Feb 26, 2026 | ||
| Knowledge-Intensive Reasoning | HQA | 18 | Feb 26, 2026 | ||
| Universal Targeted Adversarial Attack | Seen Samples (Used for Optimization) (train) | 18 | Feb 26, 2026 | ||
| Universal Targeted Adversarial Attack | Unseen (test) | 18 | Feb 26, 2026 | ||
| Vision-Intensive Perception | V* Benchmark | 18 | Feb 26, 2026 | ||
| Multimodal Reward Modeling | MM-RLHF-RewardBench | 18 | Apr 23, 2026 | ||
| Adversarial Robustness | AutoDAN | 18 | Feb 26, 2026 | ||
| Adversarial Robustness | PAIR | 18 | Feb 26, 2026 | ||
| Safety Evaluation | FalseReject | 18 | Feb 26, 2026 | ||
| Text-to-Image Generation | DrawBench | 18 | Feb 26, 2026 | ||
| Simulatability | Tabular Datasets Aggregate (test) | 18 | Feb 26, 2026 | ||
| Time Series Classification | UEA 8 multivariate | 18 | Feb 26, 2026 | ||
| Safety Evaluation | JailbreakV-28K v1 (test) | 18 | Feb 26, 2026 | ||
| Software engineering | SWE-Bench Verified | 18 | Apr 16, 2026 | ||
| Mathematical Reasoning | MATH (test) | 18 | Feb 26, 2026 | ||
| Hyper-Relational Link Prediction | JFFI V1 | 18 | Feb 26, 2026 | ||
| Customized few-shot classification | Cards | 18 | Feb 26, 2026 | ||
| Customized few-shot classification | Clevr 4 10k | 18 | Feb 26, 2026 | ||
| Web Browsing Competition (Chinese) | Browse Comp ZH | 18 | May 13, 2026 | ||
| Open-Ended Professional Tasks | BizBench 1.0 (Leaderboard) | 18 | Feb 26, 2026 |