Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1988 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Text Classification | Ohsumed | 17 | Jun 24, 2026 | ||
| Visual Mathematical Reasoning | DynaMath | 17 | Jun 23, 2026 | ||
| Text-to-Speech | Seed-TTS-Eval (test-hard) | 17 | Jun 23, 2026 | ||
| Image Denoising | SIDD Raw | 17 | Jun 23, 2026 | ||
| Fairness Evaluation | German Credit Dataset | 17 | Jun 23, 2026 | ||
| Medical Image Segmentation | Hecktor 2022 | 17 | Jun 23, 2026 | ||
| Medical Image Segmentation | AutoPET-II | 17 | Jun 23, 2026 | ||
| Brain Tumor Segmentation | BraTS 2021 | 17 | Jun 23, 2026 | ||
| Medical Image Segmentation | Hecktor 2022 | 17 | Jun 23, 2026 | ||
| Medical Image Segmentation | AutoPET-II | 17 | Jun 23, 2026 | ||
| Multi-Agent Pathfinding | Room map | 17 | Jun 23, 2026 | ||
| Multi-Agent Pathfinding | Gap map | 17 | Jun 23, 2026 | ||
| Visual Grounding | 3D FORCE REF 1.0 (test) | 17 | Jun 23, 2026 | ||
| Spatial Reasoning | 3D FORCE SAG 1.0 (test) | 17 | Jun 23, 2026 | ||
| Out-of-Distribution Detection | BDD OpenImages ID OOD | 17 | Jun 23, 2026 | ||
| Out-of-Distribution Detection | BDD ID MS-COCO OOD | 17 | Jun 23, 2026 | ||
| Out-of-Distribution Detection | PASCAL VOC ID OpenImages OOD | 17 | Jun 23, 2026 | ||
| Log-income regression (coefficient estimation) | US Census Data (whole population) | 17 | Jun 23, 2026 | ||
| Regression coefficient estimation | Politeness dataset 5512 online requests (full) | 17 | Jun 23, 2026 | ||
| Language Modeling | Wikitext-2 v1 (test) | 17 | Jun 23, 2026 | ||
| Backlit Image Enhancement | Backlit300 (test) | 17 | Jun 23, 2026 | ||
| Backlit Image Enhancement | BAID (test) | 17 | Jun 23, 2026 | ||
| Generalized Category Discovery | Aircraft | 17 | Jun 23, 2026 | ||
| German passivization | German passivization (generalization set) | 17 | Jun 23, 2026 | ||
| German passivization | German passivization (test) | 17 | Jun 23, 2026 |