Struct ChatCompletionCreateParams 

pub struct ChatCompletionCreateParams {

    pub messages: Vec<Message>,
    pub provider: Option<Provider>,
    pub model: String,
    pub frequency_penalty: Option<f64>,
    pub logit_bias: Option<IndexMap<String, i64>>,
    pub max_completion_tokens: Option<u64>,
    pub presence_penalty: Option<f64>,
    pub stop: Option<Stop>,
    pub temperature: Option<f64>,
    pub top_p: Option<f64>,
    pub max_tokens: Option<u64>,
    pub min_p: Option<f64>,
    pub reasoning: Option<Reasoning>,
    pub repetition_penalty: Option<f64>,
    pub top_a: Option<f64>,
    pub top_k: Option<u64>,
    pub verbosity: Option<Verbosity>,
    pub logprobs: Option<bool>,
    pub top_logprobs: Option<u64>,
    pub response_format: Option<ResponseFormat>,
    pub seed: Option<i64>,
    pub tool_choice: Option<ToolChoice>,
    pub tools: Option<Vec<Tool>>,
    pub parallel_tool_calls: Option<bool>,
    pub prediction: Option<Prediction>,
    pub stream: bool,
    pub stream_options: StreamOptions,
    pub usage: Usage,
}

Chat completion request parameters formatted for the OpenRouter API.

Combines parameters from both the Ensemble LLM configuration and the incoming request to create a complete request for OpenRouter.

Fields

messages: Vec<Message>

Messages for the conversation, including any prefix/suffix from the Ensemble LLM.

provider: Option<Provider>

Provider preferences merged from request and Ensemble LLM.

model: String

The model identifier from the Ensemble LLM.

frequency_penalty: Option<f64>

Frequency penalty from Ensemble LLM.

logit_bias: Option<IndexMap<String, i64>>

Logit bias from Ensemble LLM.

max_completion_tokens: Option<u64>

Maximum completion tokens from Ensemble LLM.

presence_penalty: Option<f64>

Presence penalty from Ensemble LLM.

stop: Option<Stop>

Stop sequences from Ensemble LLM.

temperature: Option<f64>

Temperature from Ensemble LLM.

top_p: Option<f64>

Top-p (nucleus sampling) from Ensemble LLM.

max_tokens: Option<u64>

Maximum tokens (legacy) from Ensemble LLM.

min_p: Option<f64>

Min-p sampling from Ensemble LLM.

reasoning: Option<Reasoning>

Reasoning configuration from Ensemble LLM.

repetition_penalty: Option<f64>

Repetition penalty from Ensemble LLM.

top_a: Option<f64>

Top-a sampling from Ensemble LLM.

top_k: Option<u64>

Top-k sampling from Ensemble LLM.

verbosity: Option<Verbosity>

Verbosity setting from Ensemble LLM.

logprobs: Option<bool>

Whether to include log probabilities from request.

top_logprobs: Option<u64>

Number of top log probabilities to return from request.

response_format: Option<ResponseFormat>

Response format specification from request.

seed: Option<i64>

Random seed from request.

tool_choice: Option<ToolChoice>

Tool choice configuration from request.

tools: Option<Vec<Tool>>

Available tools from request.

parallel_tool_calls: Option<bool>

Whether to allow parallel tool calls from request.

prediction: Option<Prediction>

Prediction hints from request.

stream: bool

Always true for streaming requests.

stream_options: StreamOptions

Stream options for usage inclusion.

usage: Usage

Usage reporting options.

Implementations

impl ChatCompletionCreateParams

pub fn new_for_chat( ensemble_llm: &EnsembleLlm, request: &ChatCompletionCreateParams, ) -> Self

Creates request parameters for a chat completion.

Applies the Ensemble LLM’s prefix/suffix messages and decoding parameters.

pub fn new_for_vector( vector_pfx_indices: &[(String, usize)], ensemble_llm: &EnsembleLlm, request: &VectorCompletionCreateParams, ) -> Self

Creates request parameters for a vector completion vote.

Transforms the vector completion request into a chat completion that asks the LLM to select from labeled response options.

Trait Implementations

impl Clone for ChatCompletionCreateParams

fn clone(&self) -> ChatCompletionCreateParams

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ChatCompletionCreateParams

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ChatCompletionCreateParams

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ChatCompletionCreateParams

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

where __S: Serializer,

Serialize this value into the given Serde serializer.