Struct LLMRequest 

pub struct LLMRequest {
    pub system: Option<String>,
    pub messages: Vec<ChatMessage>,
    pub temperature: f32,
    pub max_tokens: Option<u32>,
    pub model: Option<String>,
    pub response_format: Option<ResponseFormat>,
    pub tools: Option<Vec<ToolDefinition>>,
}

A chat completion request to an LLM provider.

This struct implements Default so callers can use struct-update syntax to stay forward-compatible with future field additions:

let req = LLMRequest {
    system: Some("...".into()),
    messages: vec![ChatMessage::user("hi")],
    ..LLMRequest::default()
};

New fields added to LLMRequest in future non-breaking releases are absorbed by ..LLMRequest::default() and will not break such call sites (unlike full struct literals, which must enumerate every field). For the fluent equivalent, see LLMRequest::builder.

Fields

system: Option<String>

Optional system prompt prepended to the conversation.

messages: Vec<ChatMessage>

Ordered list of chat messages forming the conversation.

temperature: f32

Sampling temperature (0.0–2.0).

max_tokens: Option<u32>

Maximum tokens to generate. None uses the provider default.

model: Option<String>

Model override for this request. None uses the client default.

response_format: Option<ResponseFormat>

Desired response format. None uses the provider default.

Forwarded to the provider by OpenAIClient (OpenAI response_format) and, for ResponseFormat::JsonSchema, by AnthropicClient (Anthropic output_config.format). ResponseFormat::Json without a schema has no native Anthropic equivalent and is a no-op there.

tools: Option<Vec<ToolDefinition>>

Tool definitions available to the model for this request.

Forwarded to both OpenAI (tools with type: "function") and Anthropic (tools with input_schema). Any tool calls the model makes are returned in LLMResponse::tool_calls.

Implementations

impl LLMRequest

pub fn builder() -> LLMRequestBuilder

Create a new builder for constructing an LLMRequest.

pub fn into_openai_messages(self) -> Vec<(String, String)>

Convert into OpenAI-format messages, consuming the request.

Prepends a system message if self.system is set.

pub fn into_anthropic_messages(self) -> Vec<(String, String)>

Convert into Anthropic-format messages, consuming the request.

Returns only user/assistant messages (system is handled separately by Anthropic API).

Trait Implementations

impl Clone for LLMRequest

fn clone(&self) -> LLMRequest

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LLMRequest

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

Formats the value using the given formatter.

impl Default for LLMRequest

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for LLMRequest

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for LLMRequest

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.