mpai
Join a teammate's Codex or Claude Code session from your own terminal — with your identity intact.
| What is it | Join a teammate's Codex or Claude Code session from your own terminal — with your identity intact. |
|---|---|
| Pricing | Free |
| Free tier | Yes |
| Platform | Web Application |
| Best for | pair programming with AI agents, sharing context-heavy coding sessions with teammates |
| Domain registered | 2013 |
Data updated Aug. 4, 2026
What does mpai do?
mpai is an open-source tool that turns single-user AI coding sessions into shared, multiplayer experiences. If your teammate is deep in a Codex or Claude Code conversation on their Mac, you can jump into that exact session from your own terminal — no screen sharing, no transcript pasting, no switching to a new IDE. Your prompts land with your name attached, and the host keeps full control over what's shared.
The tool works as a lightweight coordination layer above existing AI agents. The host installs mpai via Homebrew or npm, picks a running session, and generates a private invite command. A teammate with Node.js 20+ pastes that command and enters the session immediately — no global install, no separate account. mpai uses Tailscale for a private device network, so no public relay is involved. Standout features include live transcript streaming, session switching, role-based access (viewer vs. participant), and an append-only audit trail of every prompt.
mpai is ideal for small engineering teams already using Codex or Claude Code who want to collaborate in real time without losing context. It's especially useful for pair programming, debugging together, or onboarding a teammate into a complex session. The alpha currently supports macOS and requires an existing Tailscale network, but the MIT license and open-source code make it easy to inspect and adapt.
Key features
What makes it stand outWho is mpai for?
Who benefits most from this toolTrust & presence
Alternatives in Developer Tools
Visual AI pair programming tool — collaborate with your team and multiple LLMs to generate and evaluate code through a node-based interface.
AI-powered terminal with built-in agents that write code, debug issues, and deploy software directly from command line
AI coding agent that learns your coding style — ships, fixes, tests, and refactors code automatically
Give AI coding agents safe, read-only access to your production database, logs, and analytics for better debugging.
Open source AI coding agent that works in your terminal, IDE, or desktop app.
Local AI command center — build apps, run coding agents, and connect 70+ AI providers from your own machine
CLI tool that analyzes your AI coding sessions to generate performance reports and improvement insights
We provide an OpenAI-compatible orchestration layer that lets teams compose their own “virtual models” on top of any LLM - combining prompts, reasoning, review, and guardrails, and use them everywhere—from your IDE to your backend. Key features 🔌 One OpenAI-compatible API for many LLMs 🧱 Custom models you name & reuse 🧠 Reasoning mode on demand ✅ Built-in review mode 🛡️ Guardrails & PII masking 🧑💻 IDE & CLI integrations 📊 Analytics & cost controls