Crate walrus_model 

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

Merges all provider backends (OpenAI, Claude, Local) with the Provider enum, ProviderManager, and ProviderConfig into a single crate. Config uses ApiStandard (OpenAI or Anthropic) to select the API protocol. All OpenAI-compatible providers route through the OpenAI backend.

Re-exports

pub use config::HfModelConfig;

pub use config::ModelConfig;

pub use manager::ProviderManager;

Modules

config

Provider configuration.

manager

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

remote

remove providers

Structs

Client

An asynchronous Client to make Requests with.

ProviderConfig

Remote provider configuration.

Enums

ApiStandard

API protocol standard for remote providers.

Provider

Unified LLM provider enum.

Functions

build_provider

Construct a remote Provider from config and a shared HTTP client.

default_model

Default model name when none is configured.