Claude 3.7 Sonnet
Anthropic·
LLMsmidcloud
Overview
Hybrid model with extended thinking for complex reasoning tasks.
Capabilities and innovations
Extended thinkingEnhanced codingAgentic reliabilityExtended Thinking ModeHybrid ReasoningImproved Tool Use
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 84.8% | unsourced |
| Humanity’s Last Exam (no tools) | 26.7% | unsourced |
| SWE-bench Verified | 56.2% | unsourced |
| MMLU | 90.7% | unsourced |
| MMLU-Pro | 82.73% | unsourced |
- SWE-bench Pro: benchmark did not exist at release
- Terminal-Bench 2.x: benchmark did not exist at release
- MMMU-Pro: benchmark did not exist at release
Links
More from Anthropic
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.