Struct OpenAiCompatibleProvider 

pub struct OpenAiCompatibleProvider { /* private fields */ }

A provider that speaks the OpenAI-compatible chat completions API. Used by: Venice, Vercel AI Gateway, Cloudflare AI Gateway, Moonshot, Synthetic, OpenCode Zen, Z.AI, GLM, MiniMax, Bedrock, Qianfan, Groq, Mistral, xAI, etc.

Implementations

impl OpenAiCompatibleProvider

pub fn new( name: &str, base_url: &str, api_key: Option<&str>, auth_style: AuthStyle, ) -> Self

pub fn new_no_responses_fallback( name: &str, base_url: &str, api_key: Option<&str>, auth_style: AuthStyle, ) -> Self

Same as new but skips the /v1/responses fallback on 404. Use for providers (e.g. GLM) that only support chat completions.

Trait Implementations

impl ModelProvider for OpenAiCompatibleProvider

fn chat<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, request: ChatRequest<'life1>, model: &'life2 str, temperature: f64, ) -> Pin<Box<dyn Future<Output = Result<ChatResponse>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Structured chat API — the single required method.

fn context_window(&self, model: &str) -> Option<usize>

Context window size in tokens for the given model.

fn supports_native_tools(&self) -> bool

Whether provider supports native tool calls over API.

fn supports_developer_role(&self, model: &str) -> bool

Whether the given model supports the developer message role (OpenAI-spec). When true, app-owned instructions are sent as a developer message. When false, they are folded into the provider’s system-equivalent role.

fn chat_stream<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, request: ChatRequest<'life1>, model: &'life2 str, temperature: f64, events: UnboundedSender<ProviderStreamEvent>, ) -> Pin<Box<dyn Future<Output = Result<ChatResponse>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Optional streaming chat API.

fn native_capabilities(&self) -> Option<ProviderNativeCapabilities>

Provider-native capabilities outside the chat/tool turn loop.

fn submit_media<'life0, 'async_trait>( &'life0 self, request: NativeMediaRequest, ) -> Pin<Box<dyn Future<Output = Result<NativeMediaResponse>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Submit a provider-native media operation.

fn poll_media_job<'life0, 'life1, 'async_trait>( &'life0 self, job: &'life1 NativeMediaJob, ) -> Pin<Box<dyn Future<Output = Result<NativeMediaResponse>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Poll an async provider-native media job.

fn warmup<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Warm up the HTTP connection pool (TLS handshake, DNS, HTTP/2 setup). Default implementation is a no-op; providers with HTTP clients should override.