Source: Claude API Docs — Thinking
Anthropic's reference for Claude thinking: modes per model, display, signatures, tool use, interleaving, caching, context-window rules.
Contents
- What: official API documentation page
- Publisher: Anthropic
- Date: accessed 2026-11-14 (living document; API version header
2026-10-21in examples) - Original: Thinking
Summary
The mode-neutral reference for how Claude’s thinking works on the Messages API. It gives a per-model table of what each thinking value does (adaptive, enabled + budget_tokens, between_tools, disabled), explains thinking blocks and their encrypted signature, the display option, the rules for passing thinking blocks back during tool use, interleaved thinking, which models keep prior-turn thinking in context, prompt-caching interactions, and feature incompatibilities.
Key claims
- Thinking arrives as
thinkingcontent blocks beforetextblocks. Each carries asignature: an encrypted copy of the full reasoning that must be passed back unchanged. - Visible thinking text is never the raw chain of thought. It is a summary produced by a different model.
display: "summarized"returns it;display: "omitted"(default on Claude 5-generation models, Opus 4.7/4.8) returns an empty field. Billing is for the full thinking either way; omitting only cuts time-to-first-text-token. - On Claude 5-generation models and Claude Mythos Preview thinking is on by default as adaptive thinking; Opus 4.7/4.8 (and 4.6 and earlier) default to thinking off.
enabledwithbudget_tokensreturns 400 on Claude 4.7 and later (Mythos Preview still accepts it). Several models rejectdisabled(Opus 5.5, all Fable/Mythos); Opus 5 and Haiku 5.5 accept it only at efforthighor below. Sonnet 5.5’s lowest setting isbetween_tools(up-front thinking off), also only athighor below. effort(inoutput_config) controls how much work the whole response gets;max_tokensis the only hard ceiling on spend.- A tool-use loop is one assistant turn. Thinking blocks within that turn must be returned complete and unmodified (including
redacted_thinkingblocks); modified blocks give a 400. Toggling thinking mid-turn silently disables it for that request. - Manual extended thinking only allows
tool_choiceautoornone; adaptive thinking supports forced tool use, except on Opus 5.5, Sonnet 5.5 and Fable/Mythos 5.1 and 5.2, which reject forced tool use on every request (useautowith strict tools or structured outputs). - Interleaved thinking (reasoning between tool calls) is automatic with adaptive thinking; the thinking allocation can span the whole turn.
- Keep-all models (Opus 4.5+, Sonnet 4.6+, Haiku 5.5, Fable/Mythos) keep prior turns’ thinking in context, billed as input; older models strip it once a non-tool-result user message arrives.
- Thinking configuration and resolved effort are rendered into the prompt, so changing them invalidates cache breakpoints.
- Opus 4.7+ and Claude 5-generation models reject non-default
temperature/top_p/top_k(theirtop_pdefault is 0.99); from API version2026-10-21any sampling parameter returns a 400 on every model. Claude 4.6+ rejects assistant prefill. Max output is 128K on current models (300K via a batches beta on some). SDKs require streaming above 21,333max_tokens. - A request that tries to extract internal reasoning into the response text may be refused on recent models (
reasoning_extraction).