Mel – the agentic terminal
The AI terminal for teams.
FAQ
Questions about Mel.
Short answers to the questions people ask most about Mel, its plans and its platforms.
About Mel
- What is an agentic terminal?
- An agentic terminal is a terminal with AI agents built in: they plan the work, run commands, edit files and check the result, with you in control. Mel is an agentic terminal for macOS, Windows, Linux and VS Code.
- Is Mel an AI terminal for teams?
- Yes. On the Business plan ($50 a seat a month, from 2 seats) the whole team shares one balance of hosted-AI credits and one invoice, along with agents, rules, settings and recommended connectors. One admin pays and invites people with a link.
- Does Mel work on macOS, Windows and Linux?
- Yes. Mel is a desktop app for macOS (Apple Silicon and Intel), Windows 10 and 11, and Linux, and it also runs inside VS Code through the Mel extension.
- Is Mel an AI coding agent that runs in the terminal?
- Yes. Mel is the terminal itself, with a team of AI agents inside it that read your code, run commands, edit files and review diffs. It can also run Claude Code and Codex CLI as part of the team, each in its own pane.
- Can Mel run commands and edit files on my computer?
- Yes. The agent runs commands and edits files in your project, and configurable approvals keep you in control: you approve writes, edits and commands, or choose which ones it may run on its own.
- Can Mel review code locally?
- Yes. Mel shows every changed file as a diff in its review panel, and the agent can review a branch or a pull request for bugs and security issues, on your own machine.
- Which sign-in options does Mel have?
- You can sign in to Mel with Apple, Google or GitHub, or with an email address. Mel works with 17 connectors, each on your own account or key.
- Which AI models does Mel use?
- Mel works with Claude, GPT and Gemini, and with DeepSeek, Qwen, Kimi and GLM. Auto mode picks the right model for each task, or you can pin one. Use your own provider key for free (provider usage charges apply), or hosted credits on Pro, Max and Business.
- Is Mel open source?
- Not yet. Open source is coming soon. Mel is a free download today, and the AI agent is free on your own provider key. Paid plans (Pro $20 a month, Max $200 a month, Business $50 a seat a month) add hosted credits across Claude, GPT, Gemini, DeepSeek, Qwen and more.
- Can Mel do work beyond code?
- Yes. Besides building apps, websites, online stores, AI agents and automations, Mel can tidy the folders on your computer and turn finished work into a report you can share.
Pricing and plans
- Is the Mel terminal free?
- Yes. The full Mel terminal — blocks, palette, splits, chips, code review, editor and LSP — is a free download today on macOS, Windows and Linux. No account required.
- How do I get the AI agent?
- Sign up for a free Mel account in the app — it's instant, no waitlist. On the Free plan the agent runs on your own provider key: you pay your provider directly, at their price, with no Mel credits and no Mel markup. Upgrade to Pro for 1,500 hosted-AI credits a month if you'd rather not manage keys.
- Do I need an API key?
- On the Free plan, yes — the agent runs on a provider key you add in Settings → AI (Groq and Gemini both have free key tiers). On Pro, Max and Business you don't need one: Mel talks to a relay that holds the model keys server-side, so there's nothing to configure. You can still bring your own key on any plan, and it costs 0 Mel credits.
- Which platforms are supported?
- macOS (Apple Silicon & Intel), Windows 10/11, and Linux. It's a direct download — no account required.
- Where do my conversations live?
- Conversations are saved on your machine, under your home directory, and a copy of each one syncs to your Mel account so it follows you to another device. Turn off "Sync conversations to my account" in Settings → Privacy and nothing is uploaded.
- Can I use my own models?
- Yes — bring your own provider key any time, and let 'auto' route each task to the best one.
- What's the Business plan?
- Business is Pro for a team: $50 a seat a month, or $45 a seat a month billed yearly, with a minimum of 2 seats. Each seat adds 1,500 hosted-AI credits a month to one balance the whole team shares. One admin pays, invites people with a link, and adds or removes seats any time — one subscription, one invoice. The team also shares agents and settings: Drive items, tab configs, rules, agent profiles and recommended connectors.
- How does billing work?
- Each paid plan includes a monthly credit allotment (Pro 1,500, Max 18,000, Business 1,500 a seat, shared by the team) that works across Claude, GPT, Gemini, DeepSeek, Qwen and more. Credits reset each month; add-on packs roll over. Run out and the agent pauses — shell and your own key keep working. Manage or cancel anytime from the billing portal. Annual saves 10%.
Platforms and install
- How do I install Mel on a Mac?
- Download the disk image, open it and drag Mel to your Applications folder. One build runs on Apple Silicon and Intel, and it is signed and notarized by Apple, so it opens with no Gatekeeper warning.
- How do I install Mel on Windows?
- Run the installer. It is a per-user install with no admin prompt, a Start menu entry and an uninstall entry. ARM64 PCs have their own native build, and a portable .zip is also available.
- How do I install Mel on Linux?
- Download the tarball, extract it and run ./mel. It works on any x86_64 distribution and needs OpenGL.
- Is Mel on iPhone or iPad?
- Not yet. Mel on iPhone and iPad is coming soon. Today Mel runs on macOS, Windows, Linux and inside VS Code.
Next step
Still have a question?
Write to hello@openmel.dev. The terminal is free to download, and the agent is free on your own provider key.