Claude 4.5 Sonnet
Anthropic · August 2025
⚠ deprecatedCloseddense transformermultimodalAPI Available
Context Window200K tokens
Description
Hybrid reasoning model that blended fast responses with deep 'extended thinking' when needed. Positioned as the best coding model at its price point, with improved creativity, nuance, and reliability over Sonnet 4.
Notable Milestones
- ▸Achieved 77.2% on SWE-bench Verified at launch
- ▸Introduced checkpoints for Claude Code sessions
Benchmark Scores
SWE-benchReal-world software engineering
77.2%Key Innovations
Reasoning
ReasoningStructured step-by-step problem solving, often using chain-of-thought or tree-of-thought approaches.
Code Gen
Code GenAbility to write, debug, and understand programming code across multiple languages.
Test-Time Compute
Test-Time ComputeUsing extra computation during inference (not training) to improve answer quality — thinking longer on harder problems.