Loading documentation
Extended reasoning with an optional provider-generated summary.
Uses the OpenAI Responses API for GPT-5.6 and adaptive thinking for Claude 5. The block can return a summarized reasoning trace supplied by the provider; private hidden chain-of-thought is never exposed.
Layout: half
Options: GPT-5.6 Terra, GPT-5.6 Sol, GPT-5.6 Luna, Claude Sonnet 5
Placeholder: sk-ant-... or sk-...
Layout: half
Layout: half
Hidden by default
Layout: half
Options: Low, Medium, High
Layout: half
Placeholder: What should the model reason about? Ask a question or describe a task...
Layout: full
Placeholder: Optional instructions to guide the model's reasoning style...
Layout: full
Hidden by default
Required
Model to use for extended thinking
Required
API key for the selected provider
Required
The question or task to reason about
Optional
Optional system prompt
Optional
Legacy manual-thinking token budget
Optional
Reasoning effort: low, medium, or high
Optional
Whether to request a provider-generated reasoning summary
Primary response type:
{
"reasoningSummary": "string",
"thinkingContent": "string",
"response": "string",
"model": "string",
"tokens": "json",
"acknowledgedThought": "string"
}