Connect your AI model

Support for connecting your own AI models through MCP (the Model Context Protocol) is coming soon. It's intended to let you use models like Claude, Gemini, GPT and Mistral with Coassemble rather than being tied to a single one. This page will be updated with specifics at release.

Coassemble's approach is to be agent-friendly and not lock you to one AI provider. Through MCP, an open standard for connecting AI assistants and agents to tools and data, you'll be able to bring your own model to Coassemble and use the AI you already work with.

What this is for

Teams have different reasons for preferring one model over another: existing contracts, data handling requirements, or simply what they already build with. Connecting your own model means you can use Coassemble's AI features with the provider that suits you, and combine Coassemble with the AI tools you already use, rather than being limited to a single built-in model.

How it's intended to work

The plan is to support connecting models such as Claude, Gemini, GPT and Mistral through MCP. Because MCP is an open standard, the intent is to work across the models and tools that support it rather than a fixed list. The exact setup, supported models, and any plan requirements will be confirmed when it's released.

Frequently asked questions

Is this available now?

Not yet. Connecting your own AI model through MCP is coming soon. This page will be updated with setup steps and supported models at release.

Which models will it support?

The intent is to support models like Claude, Gemini, GPT and Mistral through MCP, rather than locking you to one. Specifics will be confirmed at release.

What is MCP?

MCP, the Model Context Protocol, is an open standard for connecting AI assistants and agents to external tools and data, so an AI tool can work with Coassemble without a one-off custom integration.