A thinking/reasoning chunk from a streaming response.

Emitted by providers that support extended thinking (Anthropic, Gemini, OpenAI o-series). Thinking chunks contain the model's internal reasoning process, separate from the final response text.

Fields

• :text — the reasoning text fragment
• :signature — optional cryptographic signature (Anthropic) for thinking block verification

Example

%LLM.Stream.Thinking{text: "Let me consider the user's question...", signature: nil}

# Anthropic includes a signature for verification
%LLM.Stream.Thinking{
  text: "The user is asking about...",
  signature: "EqQBCgIYAhIM..."
}