Technology RadarTechnology Radar

Claude Sonnet 4.6

llmcodingfrontieragentic
This item was not updated in last three versions of the Radar. Should it have appeared in one of the more recent editions, there is a good chance it remains pertinent. However, if the item dates back further, its relevance may have diminished and our current evaluation could vary. Regrettably, our capacity to consistently revisit items from past Radar editions is limited.
Adopt

Claude Sonnet 4.6, released February 17, 2026, is Anthropic's current flagship model for software engineering — combining frontier-level coding performance with a 1M token context window at the same pricing as its predecessor.

Why It's in Adopt

Claude Sonnet 4.6 is the default model in Claude Code, and early testing showed users preferred it over Sonnet 4.5 roughly 70% of the time for coding tasks. Key improvements over the previous generation:

  • Better context reading: More effectively reads and understands large amounts of code before modifying it — reducing the "AI trashes your file" problem in long sessions
  • Less code duplication: Consolidates shared logic rather than repeating it across files
  • Stronger prompt injection resistance: A major safety improvement for agentic use cases where the model is acting autonomously on your behalf
  • Computer use: Major improvements in GUI automation (clicking, typing, navigating browsers)
  • 1M token context window: Feed in entire codebases or massive documentation sets

Performance

  • Tops or ties on SWE-bench Verified (real GitHub issue resolution)
  • Matches Claude Opus 4.6 on OfficeQA (enterprise document comprehension)
  • Preferred over Opus 4.5 in Claude Code agentic workflows 59% of the time despite lower cost

How to Access It

  • Claude.ai — the default free and paid model
  • Anthropic APIclaude-sonnet-4-6 model ID
  • Amazon Bedrock and Google Vertex AI for enterprise access
  • Claude Code CLI — the default model

Key Characteristics

Property Value
Context window 1,000,000 tokens (1M, beta)
Max output 64,000 tokens
Input pricing $3 per million tokens
Output pricing $15 per million tokens
Provider Anthropic
API model ID claude-sonnet-4-6

Cost-Saving Features

  • Prompt caching: Up to 90% cost reduction on repeated context (great for long agentic sessions)
  • Batch processing: 50% cost reduction for non-real-time tasks