Claude 3.7 Sonnet
Claude 3.7 Sonnet is a hybrid reasoning model, one model that gives instant answers or visible extended thinking, with a thinking budget up to 128K tokens.
Anthropic calls it the first hybrid reasoning model on the market (company claim). Users or API callers choose standard mode or extended thinking and set a thinking budget. SWE-bench Verified 63.7% (70.3% with scaffolding). Price unchanged at $3/$15 per million tokens, thinking tokens included.
- Date
- Monday, 24 February 2025
- Lab
- Anthropic
- Kind
- model
- Access
- closed API
- Price
- $3/$15 per M tokens in/out, thinking tokens billed as output (2025-02-24)
Figures
| Measure | Value | Measured by |
|---|---|---|
| SWE-bench Verified (pass@1) | 63.7% 70.3% with custom scaffold | company |
| Max thinking budget (API) | 128K tokens | company |
Extended thinking was unavailable on the free tier. Retired from the Claude API on 2026-02-19. How it compares to OpenAI o1 and DeepSeek-R1 was not checked here.
Sources
This record was checked against its sources on 6 October 2026. How we check