Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1508 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Question Answering | SQuADShifts 1.0 (test) | 21 | Feb 26, 2026 | ||
| Summarization Evaluation | SummEval 1.0 (test) | 21 | Feb 26, 2026 | ||
| Machine Translation | MSCOCO En-Fr | 21 | Jun 16, 2026 | ||
| Machine Translation | WMT newstest 2010 (test) | 21 | Feb 26, 2026 | ||
| Document Layout Analysis | DocLayNet (test) | 21 | Apr 28, 2026 | ||
| Machine Translation | FR-EN | 21 | Feb 26, 2026 | ||
| Novel View Synthesis | Synthetic 360° scenes Blender (test) | 21 | Feb 26, 2026 | ||
| Extractive Question Answering | Molweni (test) | 21 | May 29, 2026 | ||
| Instruction Following | Vicuna benchmark zero-shot | 21 | Feb 26, 2026 | ||
| Multimodal classification | N24News (test) | 21 | Feb 26, 2026 | ||
| Semantic Segmentation | Cityscapes (val) | 21 | Feb 26, 2026 | ||
| Point cloud registration | LoKITTI | 21 | Feb 26, 2026 | ||
| Object Detection | Clipart1k 1.0 (test) | 21 | Feb 26, 2026 | ||
| Vision-and-Language Navigation | RxR seen (val) | 21 | Feb 26, 2026 | ||
| Pedestrian Detection | CityPersons 1.0 (val) | 21 | Feb 26, 2026 | ||
| Image Retrieval | GLD v2 | 21 | Mar 31, 2026 | ||
| Visual Localization | Cambridge Landmarks ShopFacade | 21 | Apr 8, 2026 | ||
| Image Generation | ImageNet 64x64 (train) | 21 | Apr 27, 2026 | ||
| Video Anomaly Detection | Avenue classic (test) | 21 | Feb 26, 2026 | ||
| Generalized Category Discovery | ImageNet-100 (test) | 21 | Feb 26, 2026 | ||
| 3D Face Reconstruction | REALY | 21 | Jun 29, 2026 | ||
| Text-to-image generation | MS-COCO zero-shot FID-30k 256x256 | 21 | Feb 26, 2026 | ||
| Video Action Classification | K400 time-correlated (val) | 21 | Feb 26, 2026 | ||
| Video Action Classification | SSv2 time-correlated (val) | 21 | Feb 26, 2026 | ||
| Video Action Classification | UCF101 time-correlated (val) | 21 | Feb 26, 2026 |