Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1346 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Equilibrium State Estimation | COVID-19 data Horizon=10 (test) | 22 | Jun 12, 2026 | ||
| DeepSearch Question Answering | Humanity's Last Exam | 22 | Jun 11, 2026 | ||
| Robot Manipulation | ManiSkill PullCube v1 | 22 | Jul 1, 2026 | ||
| Machine Translation | Kupang Malay | 22 | Jun 11, 2026 | ||
| Deep Search | BrowseComp-ZH v1.0 (test) | 22 | Jun 11, 2026 | ||
| Deep Search | XBench-DS v1.0 (test) | 22 | Jun 11, 2026 | ||
| Tactile Property Prediction and Commonsense Reasoning | VTV-150K | 22 | Jul 7, 2026 | ||
| Prompt inversion attack defense | Skytrax | 22 | Jun 11, 2026 | ||
| Object Hallucination Evaluation | POPE | 22 | Jun 24, 2026 | ||
| Flexible Job Shop Scheduling | FJSP GA 10 x 5, 15 x 10, 20 x 10 (evaluation sets) | 22 | Jun 11, 2026 | ||
| Flexible Job Shop Scheduling | FJSP PDR 10 x 5, 15 x 10, 20 x 10 (evaluation sets) | 22 | Jun 11, 2026 | ||
| Protein Property Prediction | 21 protein landscapes Extrapolation 512 training points | 22 | Jun 11, 2026 | ||
| Protein Property Prediction | 21 protein landscapes 128 training points (extrapolation) | 22 | Jun 11, 2026 | ||
| Protein Property Prediction | 21 protein landscapes 1536 train points (cross-val) | 22 | Jun 11, 2026 | ||
| Closed-loop planning | nuPlan 14-hard 1.0 (test) | 22 | Jun 11, 2026 | ||
| Closed-loop planning | nuPlan 1.0 (test14) | 22 | Jun 11, 2026 | ||
| AI-Generated Image Detection | Average Original | 22 | Jun 11, 2026 | ||
| IPMN Classification | Cyst-X External (val) | 22 | Jun 11, 2026 | ||
| IPMN Classification | Cyst-X Internal (val) | 22 | Jun 11, 2026 | ||
| Uncertainty Estimation | Average Across all subsets | 22 | Jun 9, 2026 | ||
| Emotional Content Detection | ChLGBT Emotional (test) | 22 | Jun 9, 2026 | ||
| Explicit Content Detection | ChLGBT Explicit (test) | 22 | Jun 9, 2026 | ||
| Long-video understanding | VideoMME Long (3~120 min) | 22 | Jun 9, 2026 | ||
| Tray Retrieval | held-out N=50 randomized configurations (test) | 22 | Jun 9, 2026 | ||
| Book Passing | N=50 randomized configurations (test) | 22 | Jun 9, 2026 |