Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1342 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| LLM Decode Throughput | Apple M3 base tg128 (test) | 22 | Jul 2, 2026 | ||
| Text-rich Visual Question Answering | TextVQA | 22 | Jul 1, 2026 | ||
| Hybrid SQL Querying | SemBench wildlife | 22 | Jul 1, 2026 | ||
| Database Query Answering | SemBench movie scenario v1 | 22 | Jul 1, 2026 | ||
| 3D Translational and Angular Velocity Estimation | 10 Golf Shots real-world 1.0 (test) | 22 | Jul 1, 2026 | ||
| Trigger inversion | Yahoo | 22 | Jul 1, 2026 | ||
| Regression | HEALTH (test) | 22 | Jul 1, 2026 | ||
| Classification | HEALTH (test) | 22 | Jul 1, 2026 | ||
| Regression | F&B (test) | 22 | Jul 1, 2026 | ||
| Classification | F&B (test) | 22 | Jul 1, 2026 | ||
| Multi-modal Classification | MM-IMDb (test) | 22 | Jul 1, 2026 | ||
| Slide-level classification | CAMELYON 17 (test) | 22 | Jun 30, 2026 | ||
| Slide-level classification | PANDA held-out (test) | 22 | Jun 30, 2026 | ||
| Slide-level calibration | CAMELYON17 (test) | 22 | Jun 30, 2026 | ||
| Slide-level calibration | CAMELYON16 (test) | 22 | Jun 30, 2026 | ||
| Slide-level calibration | PANDA (test) | 22 | Jun 30, 2026 | ||
| General Reasoning | MMLU-Pro | 22 | Jul 2, 2026 | ||
| Kernel Fuzzing | Linux Kernel 10 Targets | 22 | Jun 30, 2026 | ||
| Content Moderation | Dynamic Evaluation Sets Adversarial 1.0 | 22 | Jun 29, 2026 | ||
| Content Moderation and Data Security | in-house business benchmarks | 22 | Jun 29, 2026 | ||
| Domain Capability Evaluation | in-house domain capability benchmarks | 22 | Jun 29, 2026 | ||
| Document-level classification | Social Media Posts | 22 | Jun 26, 2026 | ||
| Commonsense Reasoning | WinoGAViL | 22 | Jun 26, 2026 | ||
| Traffic Generation | 24-hour traffic targets (10 detectors) | 22 | Jun 26, 2026 | ||
| Fine-Grained Visual Classification | FGVC-Aircrafts | 22 | Jun 25, 2026 |