Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1796 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Agent Task Execution | QWENCLAWBENCH | 18 | Jun 16, 2026 | ||
| AI-generated video detection | GenVidBench 1.0 (test) | 18 | Jun 16, 2026 | ||
| Comprehensive Data Intelligence | Aggregated Benchmarks (DABench, DSBench, TableBench, DABStep-R) | 18 | Jun 16, 2026 | ||
| Data Reasoning | DABStep-R | 18 | Jun 16, 2026 | ||
| Table Understanding | TableBench | 18 | Jun 16, 2026 | ||
| Data Science | DSBench | 18 | Jun 16, 2026 | ||
| Function-name selection | BFCL multiple live (300 tasks model) | 18 | Jun 16, 2026 | ||
| General Reasoning | GSM8K, MATH-500, GPQA-Diamond, TheoremQA Aggregate | 18 | Jun 16, 2026 | ||
| Theorem-oriented Reasoning | TheoremQA | 18 | Jun 16, 2026 | ||
| Scientific Reasoning | GPQA Diamond | 18 | Jun 16, 2026 | ||
| Multi-turn Tool-use | FTRL, BFCL, and ToolHop Combined | 18 | Jun 16, 2026 | ||
| Multi-turn Tool-use | BFCL | 18 | Jun 16, 2026 | ||
| Multi-turn Tool-use | FTRL | 18 | Jun 16, 2026 | ||
| Mathematics Reasoning | AIME 2025 | 18 | Jun 16, 2026 | ||
| Speaker Verification | VoxPhrase Eval-2 | 18 | Jun 16, 2026 | ||
| Speaker Verification | VoxPhrase Eval-1 | 18 | Jun 16, 2026 | ||
| Mathematical Reasoning | Minerva | 18 | Jun 16, 2026 | ||
| Code Verification | MBPP (test) | 18 | Jun 16, 2026 | ||
| Code Generation | HumanEval+ latest (held-out) | 18 | Jun 16, 2026 | ||
| Code Generation | LiveCodeBench v5 (held-out) | 18 | Jun 16, 2026 | ||
| Mathematical Reasoning | MATH-500 latest (held-out) | 18 | Jun 16, 2026 | ||
| Scanpath Generation | COCO-Search18 Target Absent | 18 | Jun 16, 2026 | ||
| Biological Relation Identification (EP) | BioREDirect | 18 | Jun 16, 2026 | ||
| Systematic Review Stopping | 135 DTA review outcomes (CLEF datasets) (aggregated) | 18 | Jun 16, 2026 | ||
| Textual long-horizon complex reasoning | BrowseComp | 18 | Jun 16, 2026 |