Why pay $20/mo for fitness AI when you can connect your own via MCP?

Most fitness apps today lock basic AI workout suggestions behind heavy monthly subscriptions while storing your health data on cloud servers.

When building OVERLOAD, we took a different approach: Bring Your Own AI. By making the app local-first and embedding a Model Context Protocol (MCP) server, users can connect Claude, Cursor, or local LLMs directly to their workout data without extra subscription fees or privacy trade-offs.

Do you prefer paying for all-in-one AI subscriptions, or using your existing AI setup with open protocols like MCP? Would love to hear how you manage your health & fitness data!