Enum Provider 

pub enum Provider {
    OpenAI(OpenAI),
    Claude(Claude),
}

Unified LLM provider enum.

The gateway constructs the appropriate variant based on ApiStandard from the provider config. The runtime is monomorphized on Provider.

Variants

OpenAI(OpenAI)

OpenAI-compatible API (covers OpenAI, DeepSeek, Grok, Qwen, Kimi, Ollama).

Claude(Claude)

Anthropic Messages API.

Implementations

impl Provider

pub fn context_length(&self, _model: &str) -> Option<usize>

Query the context length for a given model ID.

Local providers delegate to mistralrs; remote providers return None (callers fall back to the static map in wcore::model::default_context_limit).

pub async fn wait_until_ready(&mut self) -> Result<()>

Wait until the provider is ready.

No-op for remote providers. For local providers, blocks until the model finishes loading.

Trait Implementations

impl Clone for Provider

fn clone(&self) -> Provider

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Model for Provider

async fn send(&self, request: &Request) -> Result<Response>

Send a chat completion request.

fn stream( &self, request: Request, ) -> impl Stream<Item = Result<StreamChunk>> + Send

Stream a chat completion response.

fn context_limit(&self, model: &str) -> usize

Resolve the context limit for a model name.

fn active_model(&self) -> CompactString

Get the active/default model name.