Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1443 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| GUI Grounding | CUActSpot | 21 | May 13, 2026 | ||
| Complex Infographic Generation | BizGenEval easy (test) | 21 | May 13, 2026 | ||
| Complex Infographic Generation | BizGenEval hard (test) | 21 | May 13, 2026 | ||
| Confidence alignment | LogiQA | 21 | May 13, 2026 | ||
| Confidence Calibration | ReClor (test) | 21 | May 13, 2026 | ||
| Multi-task Language Understanding | MMLU | 21 | May 13, 2026 | ||
| Hallucination Evaluation | POPE Overall | 21 | May 15, 2026 | ||
| Long-Range Interaction Modeling | LRIM-16 hard | 21 | May 19, 2026 | ||
| Multimodal Perception and Reasoning | MME-RealWorld Lite | 21 | Jun 24, 2026 | ||
| Multi-hop QA Retrieval | MuSiQue | 21 | May 13, 2026 | ||
| Camera Pose Estimation | DyCheck | 21 | Jun 30, 2026 | ||
| Multimodal Reasoning | Geo3K | 21 | Jun 23, 2026 | ||
| Tool Use | RobustBench-TC Perturbed 1.0 (test) | 21 | May 13, 2026 | ||
| Tool Use | RobustBench-TC Clean 1.0 (test) | 21 | May 13, 2026 | ||
| Depth Estimation | NYU v2 | 21 | Jun 1, 2026 | ||
| Sequential Recommendation | Amazon Fashion (Overall) | 21 | May 13, 2026 | ||
| Sequential Recommendation | Steam Head Item | 21 | May 13, 2026 | ||
| Sequential Recommendation | Steam Tail Item | 21 | May 13, 2026 | ||
| Depth Estimation | ETH3D | 21 | May 15, 2026 | ||
| Autonomous Agent Task Execution under Injection | ClawBench Overall | 21 | May 13, 2026 | ||
| Autonomous Agent Task Execution under Injection | ClawBench (dev) | 21 | May 13, 2026 | ||
| Autonomous Agent Task Execution under Injection | ClawBench Office | 21 | May 13, 2026 | ||
| Autonomous Agent Task Execution under Injection | ClawBench Daily Life | 21 | May 13, 2026 | ||
| Autonomous Agent Task Execution under Injection | ClawBench Social | 21 | May 13, 2026 | ||
| Major Depressive Disorder Detection | MDD | 21 | May 13, 2026 |