Anthropic Launches Sonnet 4.6 AI Model

Anthropic has released Sonnet 4.6, a new AI model featuring a one-million-token context window and significant improvements in coding and agentic workflows. The model is being offered at a dramatically lower price point, which is expected to reshape the economics of AI agent development. The release intensifies competition in the AI model space, with Grok 4.2 also entering public beta with a multi-agent debate system.

- The one-million-token context window, now in beta for Sonnet 4.6, was a feature previously reserved for Anthropic's more expensive, flagship "Opus" models. This allows the model to process and analyze entire codebases or dozens of research papers in a single prompt. - Sonnet 4.6 is priced at $3 per million input tokens and $15 per million output tokens, the same as its predecessor, Sonnet 4.5. However, it demonstrates performance on par with or even exceeding the more expensive Opus 4.6 model on certain tasks, particularly those involving office and financial analysis. - Early access developers preferred Sonnet 4.6 over its predecessor, Sonnet 4.5, approximately 70% of the time for coding tasks. They also favored it over the previous flagship model, Opus 4.5, in 59% of cases, noting fewer hallucinations and better instruction following. - The model shows significant improvement in its ability to use computer applications. On the OSWorld benchmark, which tests an AI's capability to interact with software like Chrome and VS Code, Sonnet 4.6 scored 72.5%. - This release is part of a rapid iteration cycle for Anthropic, with Sonnet 4.6 launching just 12 days after the company's premium model, Claude Opus 4.6, was released on February 5, 2026. - Concurrent with this release, xAI's Grok 4.2 has also entered public beta, featuring a multi-agent system where four specialized AI agents debate and cross-check responses to reduce errors. - Grok 4.2 is being positioned as a tool for high-level expert tasks in medicine and engineering and features a "rapid learning" architecture that allows for weekly updates based on user feedback. - Sonnet 4.6 is now the default model for both free and pro users on claude.ai and is also available through the Claude API and on major cloud platforms like Amazon Bedrock and Google Cloud's Vertex AI.

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.