Struct OpenAICompatibleConfig 

pub struct OpenAICompatibleConfig {
    pub provider_name: String,
    pub api_key: String,
    pub model: String,
    pub base_url: Option<String>,
    pub organization_id: Option<String>,
    pub project_id: Option<String>,
    pub reasoning_effort: Option<ReasoningEffort>,
    pub parallel_tool_calls: bool,
}

Available on crate feature openai only.

Configuration for OpenAI-compatible providers.

Fields

provider_name: String

Provider display name used in error messages.

api_key: String

API key.

model: String

Model name.

base_url: Option<String>

Optional API base URL.

organization_id: Option<String>

Optional organization ID.

project_id: Option<String>

Optional project ID for providers that support it.

reasoning_effort: Option<ReasoningEffort>

Optional reasoning effort for OpenAI reasoning models.

parallel_tool_calls: bool

Whether to allow the model to call multiple tools in a single turn.

Implementations

impl OpenAICompatibleConfig

pub fn new( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Create config for an OpenAI-compatible provider.

pub fn with_provider_name( self, provider_name: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set provider display name used in errors.

pub fn with_base_url( self, base_url: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set a custom API base URL.

pub fn with_organization( self, organization_id: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set organization ID.

pub fn with_project( self, project_id: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set project ID.

pub fn with_reasoning_effort( self, effort: ReasoningEffort, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set reasoning effort for reasoning models.

pub fn with_parallel_tool_calls( self, parallel_tool_calls: bool, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Set whether parallel tool calls are allowed.

pub fn fireworks( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Fireworks AI preset.

Default model: accounts/fireworks/models/llama-v3p1-8b-instruct

pub fn together( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Together AI preset.

Default model: meta-llama/Llama-3.3-70B-Instruct-Turbo

pub fn mistral( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Mistral AI preset.

Default model: mistral-small-latest

pub fn perplexity( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Perplexity preset.

Default model: sonar

pub fn cerebras( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Cerebras preset.

Default model: llama-3.3-70b

pub fn sambanova( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

SambaNova preset.

Default model: Meta-Llama-3.3-70B-Instruct

pub fn xai( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

xAI (Grok) preset.

Default model: grok-3-mini

pub fn gemini( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Google Gemini (OpenAI-compatible) preset.

Targets Gemini’s OpenAI-compatibility endpoint, letting you use a Gemini API key and a Gemini model (e.g. gemini-3.5-flash) through the OpenAI Chat Completions wire format. Use a GEMINI_API_KEY for the api_key.

For native Gemini features (thinking levels, server-side tools, the Interactions API), prefer GeminiModel. This preset is for callers who want a single OpenAI-compatible code path across providers.

Default model suggestion: gemini-3.5-flash.

pub fn minimax( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

MiniMax preset.

Default model: minimax-m2.7

pub fn bytedance( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

ByteDance Doubao (Volcano Engine Ark) preset.

Default model: doubao-1-5-pro-256k

pub fn zhipu( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Zhipu AI (GLM) preset.

Default model: glm-5.1

pub fn baidu( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Baidu ERNIE (Qianfan) preset via OpenAI-compatible endpoint.

Default model: ernie-5

Note: Uses the Qianfan OpenAI-compatible endpoint. For the native Qianfan API with OAuth2 token exchange, use a dedicated client.

pub fn cohere( api_key: impl Into<String>, model: impl Into<String>, ) -> OpenAICompatibleConfig

Available on crate feature models only.

Cohere preset via OpenAI-compatible endpoint.

Default model: command-a-plus-05-2026

Note: For full Cohere features (citations, connectors, RAG), use the native Cohere API. This preset provides basic chat completions.

Trait Implementations

impl Clone for OpenAICompatibleConfig

fn clone(&self) -> OpenAICompatibleConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for OpenAICompatibleConfig

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for OpenAICompatibleConfig

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for OpenAICompatibleConfig

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.