Claude 3.7 Sonnet (20250219): benchmark results
February 19, 2025 Claude 3.7 Sonnet snapshot, tracked when sources report the dated API model. Provider: Anthropic. Released 2025-02-24. Access: API.
Aggregate Rating 1696 ± 7, rank #201 of 1656 rated models, from 951 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AIRTBench | 46.86 | Success Rate (%) | 100 |
| BrowserGym WebArena | 44.6 | Success rate (%, GenericAgent unless named) | 100 |
| ChemBench | 65.83 | Overall Score (%) | 100 |
| CyberSecEval2 Vulnerability Exploit | 56.82 | Accuracy (%) | 100 |
| EuroEval Danish NLU - Dansk | 72.62 | Micro F1 excluding MISC (%) | 100 |
| EuroEval English NLU - SST-5 | 72.13 | Matthews correlation coefficient (%) | 100 |
| EuroEval Norwegian NLU - ScaLA NB | 81.62 | Matthews correlation coefficient (%) | 100 |
| EuroEval Swedish NLU | 77.77 | NLU Average Score (%) | 100 |
| EuroEval Swedish NLU - ScaLA SV | 80.86 | Matthews correlation coefficient (%) | 100 |
| FLY-EVAL++ - Single-Step Prediction - Protocol Compliance | 94.8 | S1 D1 protocol score (%) | 100 |
| GDM InterCode CTF | 87.34 | Accuracy (%) | 100 |
| Galileo Tool Tasks - BFCL v3 Multi-Turn Base Multi-Function | 92 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=claude-3-7-sonnet-20250219 · How It Works · Data refreshed daily, snapshot 2026-10-20.