Claude Sonnet 4.6 is Anthropic's current recommended default model for most use cases, released February 17, 2026. It builds directly on Sonnet 4.5's strengths in agentic tasks and coding, adding improved performance on workplace tasks such as spreadsheets, documents, and structured data work, along with access to a 1M token context window via beta header.
Claude Sonnet 4.6 is Anthropic's current recommended default model for most use cases, released February 17, 2026. It builds directly on Sonnet 4.5's strengths in agentic tasks and coding, adding improved performance on workplace tasks such as spreadsheets, documents, and structured data work, along with access to a 1M token context window via beta header. In developer preference surveys conducted by Anthropic, Sonnet 4.6 was preferred over Sonnet 4.5 by 70% of developers and over Opus 4.5 by 59% — a striking result that suggests it hits a new price/performance sweet spot across the broadest range of tasks.
At the same $3/$15 per million token pricing as Sonnet 4.5, Sonnet 4.6 is a straightforward upgrade for the vast majority of applications. It is the model powering Claude.ai and most production integrations as of late February 2026.
claude-sonnet-4-6context-1m-2025-08-07 beta headerCoding: Continued best-in-class performance on SWE-bench Verified and real-world coding tasks. Computer use accuracy reached 94% on insurance processing benchmarks — a notable jump reflecting improved precision in structured real-world workflows.
Workplace Tasks: Meaningfully improved at document creation, spreadsheet generation and manipulation, structured data work, and business writing compared to Sonnet 4.5. This improvement is particularly notable for enterprise use cases.
Agentic Operation: Inherits Sonnet 4.5's strength in long-horizon autonomous tasks, with the added option of 1M context for tasks that require holding very large amounts of information in working memory simultaneously.
1M Context Window: Via the beta context-1m-2025-08-07 header, Sonnet 4.6 can process approximately 750,000 words in a single context — enough for entire codebases, large document collections, or very long autonomous task sessions.
Vision & Tool Use: Full multimodal support including image analysis, function calling, and tool use across all standard modalities.
For the absolute hardest reasoning and coding challenges, Opus 4.5 or 4.6 remains the stronger choice. The 1M context window is currently in beta and may have higher latency and costs than the standard 200K window. As with all current Claude models, there is no real-time internet access without explicit tool integrations.
claude-3-5-sonnet model IDs now return errors, with developers directed to migrate to Sonnet 4.6.February 26, 2026