Claude Sonnet 4.5 20250929 (Thinking): benchmark results
Provider: Anthropic. Released 2025-09-29. Access: API.
Unified ELO 1650 ± 1, rank #245 of 1761 rated models, from 7 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Vals AI MGSM | 94.33 | Accuracy (%) | 96 |
| Vals AI MedScribe | 84.1 | Accuracy (%) | 68.6 |
| MedCode | 44.13 | Score (self-reported) | 67.6 |
| Vals AI Finance Agent | 54.5 | Accuracy (%) | 67.5 |
| Vals AI CaseLaw v2 | 62.16 | Accuracy (%) | 66.9 |
| Vals AI IOI | 18.33 | Accuracy (%) | 61.3 |
| ProofBench | 19 | Score (self-reported) | 49.1 |
Interactive version: theaggregate.ai/model?slug=claude-sonnet-4-5-20250929-thinking · How It Works · Data refreshed daily, snapshot 2026-09-05.