Crate klieo_llm_ollama 

Ollama HTTP client implementing klieo_core::LlmClient.

Wraps Ollama’s /api/chat and /api/embeddings endpoints. Capability declaration (tool_calling, streaming, embeddings, max_context_tokens) is computed at construction time from the configured chat-model name; see capabilities::model_capabilities.

Example

use klieo_llm_ollama::OllamaClient;
let client = OllamaClient::new("http://localhost:11434", "qwen2.5:14b");

Limitations

• Ollama responses are JSON; binary modalities (vision, audio) are not yet supported (the vision capability is always false).
• Streaming chunks are decoded one-line-per-poll; very large lines buffer in memory until the newline.

Re-exports

pub use capabilities::model_capabilities;

pub use client::OllamaClient;

Modules

capabilities

Static lookup of klieo_core::Capabilities by Ollama model name.

client

OllamaClient — LlmClient impl over Ollama’s HTTP API.