Claude Sonnet 4.6
Anthropic·
LLMsmidcloud
Overview
Near-flagship performance at mid-tier pricing with 1M context window and advanced computer use.
Capabilities and innovations
1M context windowAdvanced computer useAdaptive thinkingFinancial analysisFlagship-tier at Mid-tier PricingARC-AGI-2 LeapOSWorld Computer Use
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 89.9% | unsourced |
| Humanity’s Last Exam (no tools) | 49% | unsourced |
| SWE-bench Verified | 79.6% | unsourced |
| MMLU-Pro | 87.34% | unsourced |
| Terminal-Bench 2.x | 57.3% | independent Terminal-Bench 2.1 57.30%, measured by Vals AI with the Terminus 2 harness (accessed 2026-09-23). Vals AI is the reference harness for this field so values compare across vendors; vendor and other third-party figures stay in the notes below. No Terminal-Bench 2.x value was recorded before. |
- SWE-bench Pro: not reported by the vendor
- MMLU: not reported by the vendor
- MMMU-Pro: not reported by the vendor
Links
More from Anthropic
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.