Skip to content
New — orchestrate a team of Mel agents from your terminal. Learn more about orchestrating Mel agents
mel
Sign inContact salesDownload

Mel – the agentic terminal
The AI terminal for teams.

All features

Everything Mel does today.

This page lists every Mel feature that ships today, grouped by where you use it. Each one links to its page or its doc.

Download the free terminalRead the docs

Terminal

A terminal built for real work

Command blocks

Each command and its output is one block you can select, copy, filter and bookmark.

Learn more

Block filtering

Live-filter a block's output, with regex and case-sensitive modes.

Learn more

Command palette

Search and run every Mel action from one fuzzy list.

Learn more

Tabs and split panes

Split right or down. Every pane is its own session.

Learn more

Vertical tabs

A sidebar with one row per pane, with names, colours and search.

Learn more

Tab configs

Reusable .toml recipes that open a tab with its layout, folders and startup commands.

Learn more

Clickable output

Paths in output are links. Cmd or Ctrl-click opens the file or folder.

Learn more

Autocomplete and history

Ghost completions from your history and files, and an Up-arrow history picker.

Learn more

Terminal chips

Pills above the prompt show your runtime versions, git state and folder.

Learn more

SSH sessions

Full SSH sessions in a pane, and the agent can work inside one.

Learn more

Themes

Eight bundled themes, light and dark, in Settings.

Learn more

Agent

An AI agent and a team of agents

Mel Agent

Ask in plain English. The agent reads your project, runs commands, edits files and checks the result.

Learn more

A team of agents

A lead agent splits a big task and runs an agent for each part in its own pane, with /orchestrate-agent.

Learn more

CLI agents in the team

Run Claude Code and Codex CLI as the team instead, with /orchestrate-cli.

Learn more

Many models

Claude, GPT, Gemini, DeepSeek, Qwen and more. Auto mode picks a model for each task, or you pin one.

Learn more

Configurable approvals

Choose which commands and edits need your yes. A step that cannot be undone always asks.

Learn more

Project memory

Notes the agent keeps per project and recalls across conversations.

Learn more

Web search

The agent can look things up on the live web during a task.

Learn more

Image attachments

Attach screenshots, designs and diagrams, and the agent sees them.

Learn more

Voice input

Dictate a prompt instead of typing it.

Learn more

Conversations

Collapse, resume, search and compact your agent conversations.

Learn more

Autonomous mode

Hands-off runs for work you trust the agent with.

Learn more

Code

Review and edit code in the same window

Code review panel

Every changed file as a diff card, against your uncommitted changes or a branch.

Learn more

Built-in code editor

Syntax highlighting, completion and save-time checks.

Learn more

Inline comments to the agent

Comment on diff lines and send them to the agent in one click.

Learn more

Language servers

Hover docs, go-to-definition and live errors from rust-analyzer, pyright and others.

Learn more

Project explorer

A file tree you can browse, rename in and hand files to the agent from.

Learn more

Global search

Search file contents across the project with ripgrep.

Learn more

Library and Drive

Keep what you reuse

Mel Drive

Saved workflows, prompts, notebooks and env sets, with a team section.

Learn more

Workflows

Commands with placeholders you fill in when you run them.

Learn more

Notebooks

Markdown notebooks with a slash menu and live preview.

Learn more

Environment variables

Named env sets with masked values, loaded into a pane in one click.

Learn more

Skills and the Library

Skills, prompts and MCP servers: Mel's own, yours and your project's.

Learn more

Everywhere you work

Connect Mel to your tools

Connectors and MCP

Connect a tool, or add an MCP server, and the agent can use it. Every call that changes something waits for your yes.

Learn more

Mel as an MCP server

Use Mel's coding agent from other apps that speak MCP.

Learn more

VS Code extension

The same agent as the Mel app, next to your code.

Learn more

Telegram approvals

Approve Mel's permission requests from Telegram. The first answer wins.

Learn more

Mac, Windows and Linux

A desktop app for macOS, Windows 10 and 11, and Linux.

Learn more

Plans

Free with your own key

Your own key, free

On the Free plan the agent runs on your own provider key, with no Mel markup.

Learn more

Pro and Max

Pro is $20 a month with 1,500 hosted-AI credits. Max is $200 a month with 18,000.

Learn more

Business

$50 a seat a month, shared credits and one invoice.

Learn more

Coming soon

Coming soon

  • Open source: coming soon
  • Mel on iPhone: coming soon
  • Mel on iPad: coming soon
  • Mel in the browser: coming soon
  • Mel Remote: coming soon

Get started

Try every feature, free.

The Mel terminal is free to download, no account needed. The AI agent is free on your own provider key — 1,500 hosted-AI credits a month on Pro.

Get startedDownload the free terminal