Crate walrus_model 

Model crate — LLM provider implementations, enum dispatch, configuration, construction, and runtime management.

Merges all provider backends (DeepSeek, OpenAI, Claude, Local) with the Provider enum, ProviderManager, and ProviderConfig into a single crate. Config uses flat ProviderConfig with model-prefix kind detection.

Re-exports

pub use config::ProviderConfig;

pub use config::ProviderKind;

pub use http::HttpProvider;

pub use manager::ProviderManager;

Modules

claude

Claude (Anthropic) LLM provider.

config

Provider configuration.

deepseek

DeepSeek LLM provider

http

Shared HTTP transport for OpenAI-compatible LLM providers.

local

Local LLM provider via mistralrs.

manager

ProviderManager — concurrent-safe named provider registry with model routing and active-provider swapping.

openai

OpenAI-compatible LLM provider.

Structs

Client

An asynchronous Client to make Requests with.

Enums

Provider

Unified LLM provider enum.

Functions

build_provider

Construct a Provider from config and a shared HTTP client.