Crate daimon_provider_local 

Locally-hosted model providers for the Daimon agent framework.

Covers every runtime you can point at a machine under your own control:

• ollama — Ollama, via its native /api/chat API
• llamacpp — llama.cpp’s llama-server, over its OpenAI-compatible API plus llama.cpp-native sampling extras
• llamars — llama-rs’s server, over its OpenAI-compatible API
• generic — any other OpenAI-compatible server (vLLM, LM Studio, llamafile, LocalAI, …)

All four are HTTP clients; none embed inference in-process.

Re-exports

pub use generic::OpenAiCompatible;

pub use generic::OpenAiCompatibleEmbedding;

pub use llamacpp::LlamaCpp;

pub use llamacpp_embed::LlamaCppEmbedding;

pub use llamars::LlamaRs;

pub use llamars::LlamaRsEmbedding;

pub use ollama::Ollama;

pub use ollama_embed::OllamaEmbedding;

Modules

generic

Generic OpenAI-compatible model provider.

llamacpp

llama.cpp model provider.

llamacpp_embed

llama.cpp embedding model provider.

llamars

llama-rs model provider.

ollama

Ollama local model provider.

ollama_embed

Ollama embeddings provider.