C

Anthropic: claude-sonnet-4-6-thinking

claude-sonnet-4-6-thinking
1M context128K outReasoningToolsVisionFilesCacheStructuredComputer usePrefill
Released Feb 17, 2026Knowledge cutoff Aug 2025Updated Aug 27, 2026

Claude Sonnet 4.6 is the mid-tier workhorse of Anthropic's 4.6 generation, bringing a 1M-token context window and 128K single-response output to the Sonnet price point. It uses the same adaptive thinking architecture as Opus 4.6, deciding reasoning depth on its own based on task complexity. It improves clearly over Sonnet 4.5 on coding, agentic workflows and long-document work while keeping the speed and cost profile the Sonnet line is known for.

Mode chatTokenizer ClaudeReasoning levels maxModeratedDeprecation Feb 2027

Pricing

Input price
$1.34/ 1M tokens55% off
Output price
$6.71/ 1M tokens55% off
Compatible endpoints anthropic, openaiVendor Anthropic

Uptime

Performance

Loading performance data...

Usage & Ranking

Loading usage...

Supported parameters

All providers = every upstream serving this model supports it. Some providers = depends on which upstream handles the request. Default = the value sent when you leave the parameter unset.

ParameterProvidersDefault
include_reasoningAll providers-
max_completion_tokensSome providers-
max_tokensAll providers-
reasoningAll providers-
reasoning_effortAll providers-
response_formatAll providers-
stopAll providers-
structured_outputsAll providers-
temperatureAll providers-
tool_choiceAll providers-
toolsAll providers-
top_kSome providers-
top_pAll providers-
verbosityAll providers-

Frequently asked questions

Similar models