AI BriefingClaude 3.7 Sonnet
Key point
Anthropic has unveiled Claude 3.7 Sonnet, a hybrid model capable of both instant responses and step-by-step reasoning.
Details
Anthropic announced Claude 3.7 Sonnet, the market's first hybrid reasoning model. This model combines general LLM capabilities that provide instant responses with a step-by-step extended thinking mode, where the thought process is visible to users, into a single model.
Key features are as follows:
- Flexible reasoning control: Users can choose between standard mode and extended thinking mode as needed, and API users can directly set the token budget for reasoning up to a maximum of 128K tokens, allowing them to balance speed and quality.
- Practical, real-world optimization: Rather than focusing on simple math or algorithmic competition problems, the model is optimized for complex tasks where LLMs are used in actual business environments.
- Enhanced coding performance: Coding and front-end development capabilities have been significantly improved, and Claude Code, an agentic coding tool that can perform engineering tasks directly in the terminal, was also unveiled alongside it.
Claude 3.7 Sonnet is available on all paid Claude plans, as well as through Amazon Bedrock and Google Cloud Vertex AI. Pricing remains the same as the previous model, at $3 per million input tokens and $15 per million output tokens.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.