Claude 3.7 Sonnet
Anthropic
Anthropic's first hybrid reasoning model, combining quick responses with optional extended thinking in a single model. Strong gains in coding and agentic tasks, and introduced configurable thinking budgets. Superseded by the Claude 4 family.
Strengths
- First Claude hybrid reasoning model - quick replies or extended thinking in one model
- Configurable thinking budget to trade latency for depth
- Large gains in coding and agentic workflows over 3.5 Sonnet
- 200K context with vision and strong tool use
Caveats
- Superseded by the Claude 4 family
- Extended thinking increases latency and output token cost
- 200K context is smaller than the 1M offered by Claude 4.6/4.7
- Vendor
- Anthropic
- Context Window
- 200,000 tokens
- Max Output
- 64,000 tokens
- License
- Proprietary
- Release Date
- 2025-02-24
- Training Cutoff
- 2024-11-30
- Modalities
- textimage
Capabilities
Vision
Audio
Video
Tool Use
Resources
Pricing
$3 per 1M input tokens, $15 per 1M output tokens (extended thinking output billed at output rate). Prompt caching and Batch API available.