Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2774 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Toxicity Evaluation | AttaQ 1402 prompts (test) | 14 | Feb 26, 2026 | ||
| LLM Routing | Average across Benchmarks (val) | 14 | Feb 26, 2026 | ||
| LLM Routing | BBEH (val) | 14 | Feb 26, 2026 | ||
| LLM Routing | MEDMCQA (val) | 14 | Feb 26, 2026 | ||
| LLM Routing | SUPERGPQA (val) | 14 | Feb 26, 2026 | ||
| LLM Routing | MMLU-PRO (val) | 14 | Feb 26, 2026 | ||
| LLM Routing | BBEH | 14 | Feb 26, 2026 | ||
| LLM Routing | MedMCQA | 14 | Feb 26, 2026 | ||
| LLM Routing | SUPERGPQA | 14 | Feb 26, 2026 | ||
| LLM Routing | MMLU-PRO | 14 | Feb 26, 2026 | ||
| LLM Routing | MMLU-PRO, SUPERGPQA, MEDMCQA, BBEH (test) | 14 | Feb 26, 2026 | ||
| Guarded Agent Evaluation | AgentHarm latest (full) | 14 | Feb 26, 2026 | ||
| Guarded Agent Evaluation | ASB latest (IPI) | 14 | Feb 26, 2026 | ||
| Guarded Agent Evaluation | ASB latest (DPI) | 14 | Feb 26, 2026 | ||
| Guarded Agent Evaluation | AgentDojo full latest | 14 | Feb 26, 2026 | ||
| Factual recall | MMLU | 14 | Feb 26, 2026 | ||
| Translation | Sand-Glass Zh-to-En (human evaluation) | 14 | Feb 26, 2026 | ||
| Patient message response drafting | SyPPM | 14 | Feb 26, 2026 | ||
| Clickbait Detection | NC-Clickbait | 14 | Feb 26, 2026 | ||
| Clickbait Detection | CD-Clickbait | 14 | Feb 26, 2026 | ||
| Clickbait Detection | DL-Clickbait | 14 | Feb 26, 2026 | ||
| Text-to-audio generation | VGGSound | 14 | Jun 12, 2026 | ||
| Commonsense Reasoning | IndoCulture (test) | 14 | Feb 26, 2026 | ||
| Question Answering | IndoCulture | 14 | Feb 26, 2026 | ||
| Pairwise classification | PMR-Synth (Total) | 14 | Feb 26, 2026 |