Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1992 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image Classification | ImageNet sub-tasks S1-S3 (10 classes) | 17 | Jun 11, 2026 | ||
| General Language Modeling | LiveBench | 17 | Jun 9, 2026 | ||
| Code Generation | LCB V2 | 17 | Jun 9, 2026 | ||
| Long-video understanding | LongVideoBench 0~60 min | 17 | Jun 9, 2026 | ||
| Demand Forecasting | SNBC industrial dataset | 17 | Jun 9, 2026 | ||
| Dementia care dialogue | Dementia care dialogue | 17 | Jun 9, 2026 | ||
| Signal Reconstruction | X-ray scattering wide-angle | 17 | Jun 9, 2026 | ||
| Audio Anti-spoofing | ASVspoof 5 (evaluation) | 17 | Jul 1, 2026 | ||
| Instruction Following | OmniCap-IF (test) | 17 | Jun 9, 2026 | ||
| Instructional Temporal Answer Grounding in Video | TutorialVQA (test) | 17 | Jun 9, 2026 | ||
| Hallucination Detection | LLM-AggreFact | 17 | Jun 9, 2026 | ||
| Math reasoning | MMLU Math | 17 | Jun 9, 2026 | ||
| fine-grained visual understanding | MMStar | 17 | Jun 9, 2026 | ||
| Multimodal Deep Information Seeking | BrowseComp-VL | 17 | Jun 9, 2026 | ||
| Long video understanding | LV-Omni-Bench | 17 | Jun 9, 2026 | ||
| Image Aesthetics Assessment | AVA | 17 | Jun 9, 2026 | ||
| Grounded Diagnostic Segmentation | DiagSeg | 17 | Jun 8, 2026 | ||
| One-to-Many Temporal Grounding | OMTG Bench (test) | 17 | Jun 5, 2026 | ||
| Text-to-video | Wan 2.1 | 17 | Jun 5, 2026 | ||
| CTI Rule Generation | GTI (test) | 17 | Jun 5, 2026 | ||
| UI/UX selection | WiserUI-Bench | 17 | Jun 5, 2026 | ||
| Safety Evaluation | Safety Benchmarks VLG-1, VLG-2, Ch3EF | 17 | Jun 5, 2026 | ||
| Context-aware time-series reasoning | CiK (Context-is-Key) | 17 | Jun 5, 2026 | ||
| Instructional Image Editing | RE-Edit | 17 | Jun 5, 2026 | ||
| Crime Density Forecasting | Crime Data LA region [33.9519, 33.9951] x [-118.2635, -118.2262] 2015 (last two weeks) | 17 | Jun 5, 2026 |