Struct ChatOptions 

pub struct ChatOptions {
    pub temperature: Option<f32>,
    pub max_tokens: Option<u32>,
    pub top_p: Option<f32>,
    pub stop: Option<Vec<String>>,
    pub system: Option<String>,
    pub model: Option<String>,
    pub cache_strategy: CacheStrategy,
    pub request_id: Option<String>,
    pub cancel: Option<CancellationToken>,
}

Chat completion options

Fields

temperature: Option<f32>

Temperature (0.0 - 1.0)

max_tokens: Option<u32>

Maximum tokens to generate

top_p: Option<f32>

Top-p sampling

stop: Option<Vec<String>>

Stop sequences

system: Option<String>

System prompt

model: Option<String>

Per-request model override.

When Some, providers MUST use this model name instead of their default. This enables per-session model switching without replacing the provider.

cache_strategy: CacheStrategy

Prompt-cache strategy. Ignored by providers without prompt caching.

request_id: Option<String>

Caller-supplied identifier used to attribute the resulting cost, tokens, and duration in telemetry. ChatAgent generates a fresh id per turn; multi-tenant deployments can stamp their own (e.g. "<tenant>-<msg_id>") so the analytics query cost_by_request(id) can pull every event for one request.

cancel: Option<CancellationToken>

Cooperative cancellation handle for stream_chat. When the token cancels, the provider stops emitting chunks at the next poll-point and drops the underlying HTTP connection. Has no effect on the non-streaming chat call (use tokio::select! against the future instead).

Not serialised — only meaningful inside one process.

Implementations

impl ChatOptions

pub fn new() -> ChatOptions

Create new chat options with defaults

pub fn temperature(self, temperature: f32) -> ChatOptions

Set temperature

pub fn max_tokens(self, max_tokens: u32) -> ChatOptions

Set max tokens

pub fn system<S>(self, system: S) -> ChatOptions

where S: Into<String>,

Set system prompt

pub fn top_p(self, top_p: f32) -> ChatOptions

Set top-p sampling

pub fn model<S>(self, model: S) -> ChatOptions

where S: Into<String>,

Override the model for this request.

pub fn cache_strategy(self, strategy: CacheStrategy) -> ChatOptions

Set the prompt-cache strategy.

pub fn request_id<S>(self, request_id: S) -> ChatOptions

where S: Into<String>,

Stamp a caller-supplied request_id so telemetry can attribute the resulting tokens / cost to this specific call.

pub fn cancel_with(self, token: CancellationToken) -> ChatOptions

Attach a cancellation token to stream_chat. Cancelling stops chunk delivery at the next poll-point and drops the upstream HTTP connection.

pub fn deterministic(max_tokens: u32) -> ChatOptions

Deterministic classification/routing (temp=0, few tokens)

pub fn factual(max_tokens: u32) -> ChatOptions

Low-temperature factual generation

pub fn creative(max_tokens: u32) -> ChatOptions

Creative generation with moderate temperature

Trait Implementations

impl Clone for ChatOptions

fn clone(&self) -> ChatOptions

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ChatOptions

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

Formats the value using the given formatter.

impl Default for ChatOptions

fn default() -> ChatOptions

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ChatOptions

fn deserialize<__D>( __deserializer: __D, ) -> Result<ChatOptions, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ChatOptions

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

where __S: Serializer,

Serialize this value into the given Serde serializer.