0 vs 3 benchmarks won
Anthropic Claude Opus 4 | Anthropic Claude Sonnet 4.5 | |
|---|---|---|
| Overview | ||
| Company | Anthropic | Anthropic |
| Release date | May 22 2025 | Sep 29 2025 |
| Model type | — | — |
| Open source | No | No |
| Specifications | ||
Parameters | — | — |
Context window | — | — |
| Benchmarks | ||
Science reasoning GPQA Diamond | 79.6% | 83.4% |
Software engineering SWE-Bench Verified | 72.5% | 77.2% |
Multimodal understanding MMMU | — | 68% |
| Timeline | ||
| Release gap | Claude Opus 4 shipped 130 days before Claude Sonnet 4.5 | |
Claude Sonnet 4.5 leads Claude Opus 4 on 3 of the tracked benchmarks (GPQA Diamond, SWE-Bench Verified, MMMU). Claude Opus 4 shipped 130 days before Claude Sonnet 4.5, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On GPQA Diamond, Claude Sonnet 4.5 scores 83.4%, 3.8 points above Claude Opus 4 at 79.6%. On SWE-Bench Verified, Claude Sonnet 4.5 scores 77.2%, 4.7 points above Claude Opus 4 at 72.5%.