Tome - Magical AI Spellbook

What is Tome?

Tome is a magical desktop app that lets anyone harness the power of LLMs and the Model Context Protocol (MCP). Available for macOS and Windows (Linux coming soon), it is designed for engineers, tinkerers, hobbyists, and anyone who wants to create their own AI‑powered spellbook without technical complexity.

How to use Tome?

Download and install Tome, then connect your preferred LLM provider—presets for Ollama, Gemini, and OpenAI are included, and you can add custom endpoints such as LM Studio. Open the MCP tab in the app and install MCP servers by pasting a server command (e.g., uvx mcp-server-fetch). Start chatting and let your MCP‑powered model perform tasks.

Key features of Tome

• Beginner‑friendly: no JSON, Docker, Python, or Node required
• Supports remote (Gemini, OpenAI) and local (Ollama, LM Studio) LLMs
• UI for installing, removing, and toggling MCP servers
• One‑click installation of thousands of MCP servers via Smithery.ai
• Customizable context windows and temperature
• Native support for tool calls and reasoning models with clear UI

Use cases of Tome

• Create a fully private, local AI assistant using Ollama and local MCP servers
• Connect to cloud models and remote MCP servers for state‑of‑the‑art capabilities
• Ask your model to fetch live data, like the top story from Hacker News
• Build custom AI workflows without managing packages or config files
• Experiment with different LLMs and tool‑calling patterns in a single app

FAQ from Tome

What platforms does Tome support?

Tome currently runs on macOS and Windows. Linux support is planned for a future release.

What LLM providers are compatible with Tome?

Tome works with remote providers (Google Gemini, OpenAI, and any OpenAI‑compatible endpoint) and local providers (Ollama, LM Studio, Cortex, and any OpenAI‑compatible local endpoint).

Is Tome free and open source?

Yes, Tome is licensed under Apache 2.0 and is free to use. The source code is available on GitHub.

Do I need programming skills to use Tome?

No. Tome is built for everyone—you connect your LLM and install MCP servers entirely through the UI, with no need to edit JSON, use Docker, or manage package managers.

Can I use Tome completely offline?

Yes. By choosing a local LLM (e.g., Ollama) and using only local MCP servers, all processing stays on your machine, keeping your data private and offline