Struct ChatResponse 

pub struct ChatResponse {
    pub text: Option<String>,
    pub tool_calls: Vec<ToolCall>,
    pub usage: Option<TokenUsage>,
    pub reasoning_content: Option<String>,
}

An LLM response that may contain text, tool calls, or both.

Fields

text: Option<String>

Text content of the response (may be empty if only tool calls).

tool_calls: Vec<ToolCall>

Tool calls requested by the LLM.

usage: Option<TokenUsage>

Token usage reported by the model_provider, if available.

reasoning_content: Option<String>

Raw reasoning/thinking content from thinking models (e.g. DeepSeek-R1, Kimi K2.5, GLM-4.7). Preserved as an opaque pass-through so it can be sent back in subsequent API requests — some model_providers reject tool-call history that omits this field.

Implementations

impl ChatResponse

pub fn has_tool_calls(&self) -> bool

True when the LLM wants to invoke at least one tool.

pub fn text_or_empty(&self) -> &str

Convenience: return text content or empty string.

pub fn is_semantically_empty_terminal(&self) -> bool

True when this response cannot make progress or complete a turn.

Reasoning content is intentionally excluded: it may need to be round-tripped to a provider, but it is not a user-visible final answer. A response containing one or more tool calls remains valid even when its text is empty.

Trait Implementations

impl Clone for ChatResponse

fn clone(&self) -> ChatResponse

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ChatResponse

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.