Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1257 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Instruction Following | MQUAKE | 24 | Feb 26, 2026 | ||
| Monocular Depth Estimation | NYU-Depth V2 (test) | 24 | Apr 28, 2026 | ||
| Machine Translation | WMT 2023 (test) | 24 | Feb 26, 2026 | ||
| Video depth estimation | PointOdyssey (val) | 24 | Mar 17, 2026 | ||
| Vision-Language Interaction Evaluation | Chatbot Arena Vision English queries Nov. 13, 2024 (leaderboard) | 24 | Feb 26, 2026 | ||
| Agent Task Performance | AgentDojo Travel | 24 | Feb 26, 2026 | ||
| Visual Geolocation | OSV-5M (test) | 24 | Jul 3, 2026 | ||
| 3D semantic occupancy prediction | SSCBench-KITTI-360 (test) | 24 | Apr 2, 2026 | ||
| Novel View Synthesis | RealEstate10K Easy | 24 | Jun 17, 2026 | ||
| Unsupervised Domain Adaptation | VisDA (val) | 24 | Feb 26, 2026 | ||
| Object Detection | AODRaw 1.0 (test) | 24 | Feb 26, 2026 | ||
| Scene Text Recognition | Uber | 24 | Feb 27, 2026 | ||
| Scene Text Recognition | COCO | 24 | Feb 27, 2026 | ||
| Scene Text Recognition | ArT | 24 | Feb 27, 2026 | ||
| Architecture performance prediction | NATS-Bench SSS ImageNet16-120 (stratified subset) | 24 | Feb 26, 2026 | ||
| Architecture performance prediction | NATS-Bench SSS CIFAR-100 (stratified subset) | 24 | Feb 26, 2026 | ||
| Architecture performance prediction | NATS-Bench SSS CIFAR-10 (stratified subset) | 24 | Feb 26, 2026 | ||
| Video Anomaly Detection | ShanghaiTech (SHTech) (test) | 24 | Feb 26, 2026 | ||
| Adversarial Attack | MNIST (test) | 24 | Feb 26, 2026 | ||
| Single-asset trading | TSLA (test) | 24 | Feb 26, 2026 | ||
| Spurious Correlation Discovery | MNIST, FashionMNIST, and COCO 654 evaluation settings (test) | 24 | Feb 26, 2026 | ||
| Continual Learning | CORe50 | 24 | Jun 12, 2026 | ||
| Embodied Question Answering | OpenEQA EM-EQA | 24 | Jul 3, 2026 | ||
| Image Denoising | Kodak Gaussian σ=25 (test) | 24 | Feb 26, 2026 | ||
| Gaussian Deblurring | ImageNet 256 x 256 (val) | 24 | May 15, 2026 |