Claude 3.7 Sonnet (Non-reasoning): benchmark results
Claude 3.7 Sonnet evaluated with reasoning disabled. Provider: Anthropic. Released 2025-02-24. Access: API.
Aggregate Rating 1586 ± 21, rank #745 of 2168 rated models, from 105 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Nejumi 4 - jaster (2-shot) - ALT English-to-Japanese | 92.67 | COMET wmt22 (x100) | 97.2 |
| Nejumi 4 - Toxicity - Violation Categories | 55.56 | Criteria met (%) | 93.4 |
| Nejumi 4 - BFCL - Non-Live AST | 80.37 | Accuracy (%) | 85.3 |
| Nejumi 4 - jaster (0-shot) - JSeM | 80 | Exact match (%) | 85 |
| Nejumi 4 - jaster (2-shot) - JSeM | 81 | Exact match (%) | 83.9 |
| Nejumi 4 - jaster (0-shot) - CommonsenseMoralJA | 96 | Exact match (%) | 82.9 |
| AA-Omniscience Index - Law | -1.9 | Omniscience Index | 79.1 |
| Nejumi 4 - jaster (0-shot) - JSQuAD | 80.44 | Character F1 (x100) | 79 |
| AA-Omniscience Hallucination Rate | 52.81 | Hallucination Rate (%) | 76.3 |
| AA-Omniscience Index - Business | -5 | Omniscience Index | 75 |
| AA-Omniscience Index - Humanities & Social Sciences | -3.6 | Omniscience Index | 73.2 |
| Nejumi 4 - Toxicity - Social Norms | 97.22 | Criteria met (%) | 72 |
Interactive version: theaggregate.ai/model?slug=claude-3-7-sonnet-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.