Struct CompletionOptions 

pub struct CompletionOptions<'a> {
    pub model: &'a str,
    pub max_tokens: Option<usize>,
    pub temperature: Option<f32>,
    pub timeout: Option<Duration>,
    pub stop_sequences: Vec<String>,
}

Options for a completion request, passed to LlmProvider::complete_with_options.

Allows callers to supply per-request parameters (max tokens, temperature, timeout) in addition to the model name, without changing the base LlmProvider::complete signature.

Fields

model: &'a str

Model identifier (e.g. "claude-sonnet-4-6", "gpt-4o").

max_tokens: Option<usize>

Maximum number of output tokens. Overrides provider defaults when set.

temperature: Option<f32>

Sampling temperature in [0.0, 2.0]. Higher = more random.

timeout: Option<Duration>

Per-request wall-clock timeout.

stop_sequences: Vec<String>

Stop sequences: the model will stop generating when it produces any of these strings. An empty slice means no stop sequences.

Implementations

impl<'a> CompletionOptions<'a>

pub fn new(model: &'a str) -> Self

Create options with just a model name and all other fields unset.

pub fn with_max_tokens(self, n: usize) -> Self

Set the maximum output tokens.

pub fn with_temperature(self, t: f32) -> Self

Set the sampling temperature.

pub fn with_timeout(self, d: Duration) -> Self

Set the per-request timeout.

pub fn with_stop_sequences(self, sequences: Vec<String>) -> Self

Set stop sequences for this request.

pub fn with_timeout_secs(self, secs: u64) -> Self

Set the per-request timeout from a number of seconds.

pub fn with_timeout_ms(self, ms: u64) -> Self

Set the per-request timeout from a number of milliseconds.

pub fn has_stop_sequences(&self) -> bool

Return true if at least one stop sequence has been configured.

pub fn stop_sequence_count(&self) -> usize

Return the number of stop sequences configured.

Trait Implementations

impl<'a> Debug for CompletionOptions<'a>

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

Formats the value using the given formatter.