I built Proxima, a multi-AI MCP server that runs locally on your machine—Proxima enables multi AI collaboration without paid API keys

Hi everyone,

I wanted to share a project I’ve been building: Proxima — an open-source multi‑AI MCP server that lets multiple AI models collaborate autonomously, without requiring separate paid API keys.

I recently recorded a session where I triggered a “hive‑mind” workflow across five models (Claude, ChatGPT, Gemini, and Perplexity) with one simple prompt:

“Build an innovative 3D website concept.”

GitHub: Zen4-bit/Proxima

What happened next

• Autonomous collaboration: The models didn’t just respond—they challenged, refined, and debugged each other’s ideas in real time.
• Direct execution: Using Proxima’s agentic workflow, the AI team ran terminal commands (e.g., npm install, vite) to bootstrap the environment.
• Result: They built Chronos Weave —a functional 3D spatial timeline with glassmorphism and physics‑based interactions—without me writing a single line of code.

Key features of Proxima

• Zero API costs: Works via your existing browser/account sessions.
• Agentic workflows: Ideal when you want multiple perspectives (architecture, UI/UX, implementation) operating together.
• Local‑first privacy: Everything runs on your own server.

I’d love feedback from the community on this approach to AI orchestration—especially what you’d want to see next (workflows, integrations, guardrails, etc.).

What do you think?