Osaurus FAQs

Question 1

Does Osaurus support tool calling and integration with AI agents?

Accepted Answer

Yes, Osaurus acts as a full MCP (Model Context Protocol) server, enabling tool calling and seamless integration with MCP clients like Cursor or Claude Desktop. Its extensible plugin system allows AI agents to utilize various official or custom tools.

Question 2

What is Osaurus and what does it do?

Accepted Answer

Osaurus is a native macOS LLM server optimized for Apple Silicon. It allows you to run large language models locally with OpenAI and Ollama-compatible APIs, supporting advanced features like tool calling and a comprehensive plugin ecosystem.

Question 3

Can I integrate Osaurus with existing OpenAI or Ollama SDKs?

Accepted Answer

Absolutely. Osaurus provides API endpoints identical to OpenAI's `/v1/chat/completions` and Ollama's formats, including streaming and tool calling. You can simply point your existing OpenAI or Ollama SDK clients to the Osaurus server's base URL.

Question 4

What platforms and hardware does Osaurus require?

Accepted Answer

Osaurus is designed specifically for macOS 15.5+ and requires an Apple Silicon (M1 or newer) processor for optimal performance. Apple Foundation Model support requires macOS 26+.

Question 5

What are the key benefits of using Osaurus for local LLMs?

Accepted Answer

Osaurus provides highly optimized inference for Apple Silicon via MLX, offers full OpenAI & Ollama API compatibility for easy integration, includes a robust plugin system for extended capabilities (browser, filesystem, git, search), and features a convenient Menu Bar Chat UI.

Osaurus

Osaurus

Key Features

Use Cases

Key Features

Use Cases