Module thinking 

Thinking/Reasoning Level Control

Structs

ResolvedThinking

ThinkingConfig

Configuration for thinking/reasoning level control.

ThinkingParams

Parameters derived from a thinking level, applied to the LLM request.

Enums

ThinkingDisplayMode

User-facing control for Anthropic’s thinking.display beta (thinking-display-updates-2026-08-18), which shapes how thinking blocks come back: omitted, as progress updates, or summarized. Off (the default) leaves the field out of requests entirely, matching pre-beta behavior.

ThinkingLevel

How deeply the model should reason for a given message.

Functions

apply_thinking_level

Convert a ThinkingLevel into concrete parameters for the LLM request.

apply_thinking_level_with_config

Convert a ThinkingLevel into parameters, resolving native extended thinking from the provided config.

clamp_temperature

Clamp a temperature value to the valid range [0.0, 2.0].

parse_thinking_directive

resolve_thinking_from_message

resolve_thinking_level

strip_thinking_directive

validate_thinking_config

Validate thinking config at startup. Call once during agent initialization to warn about unrecognized budget_tokens keys.