10x your Claude Code workflow.
Orchestrate dozens of agents. From any device.

Monet gives you a control room for every agent. Bring your own API keys, run as many Claude Code sessions as you want in parallel, read conversations in beautiful native UI, and organize everything by project — on your Mac, iPad, or iPhone.

The first Agentic IDE. A lightweight daemon on your Mac manages real Claude Code processes — same filesystem, same permissions, same tools you already use. Connect from any device over your own network.

Alpha — buggy, ambitious, shipping daily

Up to 3 concurrent agents free. Unlimited with a license.

The only Claude Code client for every device

macOS
iOS
iPadOS
Desktop
Browser
Without Monet
  • ×One agent at a time
  • ×Babysit the terminal window
  • ×Alt-tab to check if it's done
  • ×Miss tool approval prompts
  • ×Copy-paste context between sessions
  • ×Squint at raw ANSI output
With Monet
  • Dozens of agents working in parallel
  • Grid view with live status badges
  • Push notifications when agents need you
  • Approve tools with one tap
  • Organized by project with drag-and-drop
  • Readable conversation UI
Multi-Agent

Run a fleet of agents,
not a single terminal

Launch dozens of terminals running Claude Code in parallel — each working on a different task, all visible at once. One agent refactors auth, another writes tests, a third fixes bugs. Cards glow green while working, turn orange when they need help, and blue when ready for the next task.

Parallel AgentsGrid ViewLive StatusSession Persistence
Multiple Claude Code agents running in parallel
Session Management

See every agent
at a glance

Your control room for every agent. Tabs filter by project, and each card shows a live AI-generated summary of what the agent is doing right now (bring your own LLM key to summarize — it never goes to Monet servers).

Four layout densities — from full grid cards to a compact list of 20+ sessions — let you monitor an entire fleet without switching windows. Click any card to drop in.

4 LayoutsLive SummariesProject TabsQuick Switch
Grid layout showing agent sessions as cardsList layout showing agent sessions with descriptionsCompact list layout showing many sessions at once
Readable Mode

Claude Code,
readable

Every session is a real terminal — full xterm.js, same as running Claude Code locally. But when you want clarity, toggle readable mode: Monet optionally parses each agent's conversation into beautiful message cards — user prompts, assistant responses, tool calls, and file changes. Approve tools, answer questions, and review diffs without squinting at raw ANSI output.

Message CardsTool ApprovalsDiff ReviewNative UI
Readable mode showing formatted conversation
Terminal view showing raw Claude Code output
Terminal
Readable

Download Monet

Up to 3 concurrent agents included. Activate a license for unlimited.

Download for Mac

macOS 13+ · Apple Silicon

Up and running in minutes

Three steps. No accounts, no cloud setup, no configuration files.

Drag Monet to Applications to install
Step 1

Install Monet .dmg

Download the .dmg. Drag to Applications.

Runs as a background service — starts on boot.

Monet daemon running in the macOS menu bar
Step 2

Start the Daemon

Enter your hostname and token.

Desktop, iOS, or any browser — encrypted.

Connect to daemon and launch Claude Code agents
Step 3

Open Agentic IDE

Spin up Claude Code sessions in parallel.

Approve tools, monitor progress, read conversations.

Organization

One project,
many agents

Stop waiting on one agent to finish before starting the next. Organize sessions by project, assign a dozen agents to different tasks, and jump between them instantly with ⌘K. While one agent refactors your auth module, another is writing tests and a third is fixing bugs — all in the same project, all at once.

ProjectsParallel Work⌘K PaletteZero Wait Time
Command palette searching across sessions and projects
Mobile Access

Take your agents
with you

Monitor progress from your phone, approve tools from your lock screen, and get push notifications the moment any agent needs attention. Full session management from iOS — launch agents, switch projects, and review diffs without touching your laptop.

iOS AppPush NotificationsLock Screen ActionsFull Control
Watch me build a website on the chairlift →
Building sites from the chairlift — coding on iPhone with Monet
Your API Keys

Bring your own keys.
No subscription.

Every Claude Code session is a normal Claude Code session — your existing API key, your config, your permissions. Monet doesn't proxy, intercept, or modify anything. The only API key Monet needs is for optional session summaries. No subscription, no usage fees.

Normal Claude CodeNo ProxyNo Usage FeesYour Config
Settings
MONET-X4WE-F7SK-9RMN
License active — unlimited sessions
sk-ant-api03-••••••••••••••••kQ
ElevenLabs
sk-••••••••••vR
Deepgram
dg-••••••••••mK
Claude Code sessions use your existing Claude Code config. Monet doesn't touch your API keys or settings.
Local-First

Your Mac, your rules.
No approval needed.

Monet is a terminal app. Not a cloud IDE, not a sandbox, not a container — a terminal. Your agents run with the exact same permissions because it is your terminal. No IT tickets, no admin approval, no security review. Install it like any other Mac app.

100% LocalSame PermissionsNo CloudNo Approval
Runs on your Mac
A native daemon — same as running a terminal. Your filesystem, your SSH keys, your git config.
No data leaves your machine
Code stays local. Only outbound call is license validation — no telemetry, no analytics, no code scanning.
Nothing to approve
No new cloud service to vet. No permissions to grant. If you can install a .dmg, you can use Monet.
Power User

Keyboard-first
navigation

Never leave the keyboard. Every action in Monet has a shortcut — jump between sessions, switch projects, search anything. Built for the kind of developer who already uses vim motions and tmux panes.

Command PaletteGrid NavQuick SwitchKeyboard-First
K
Command palette
Search sessions, projects, actions
CtrlCtrl
Grid view
Double-tap to toggle grid overlay
1-9
Switch sessions
Jump directly to any session
T
New session
Launch a new agent
W
Close session
Close the current session
O
Next agent in need
Jump to the next session waiting for input
Esc
Back to grid
Exit session, return to overview
Voice Input

Dictate code commands,
accurately

Regular speech-to-text mangles code. Monet's code-aware dictation knows the difference between "user.rb" and "use our Arby's". Speak naturally — filenames, flags, and method names come through exactly right.

Code-Aware STTFilenamesCLI FlagsMethod Names
You say it. Monet gets it.
refactor user.rb to use concerns
Regular STT
refactor user dot are b to use concerns
Monet
refactor user.rb to use concerns
run pytest dash dash cov
Regular STT
run pie test – – cough
Monet
run pytest --cov
add a useEffect cleanup for the WebSocket
Regular STT
add a use effect clean up for the web socket
Monet
add a useEffect cleanup for the WebSocket

Alpha pricing

It's early. There are bugs. I'm fixing them as fast as I can.
You get a discount for putting up with me.

Alpha — Until March 20
License
$99$499$249/ year
per user, one-time annual purchase
Alpha pricing. Bugs included free of charge.
7-day money-back guarantee
You'll receive an invoice — almost always expensable.
  • Unlimited Claude Code sessions
  • Bring your own API keys — no usage fees
  • Multi-agent grid view
  • Readable conversation mode
  • Project organization
  • iOS, iPadOS, and desktop apps
  • Push notifications & voice input
  • Session persistence
  • All future updates (there will be many)

Need team licensing or custom integrations?

Contact us for enterprise pricing
6x
Throughput vs. single-agent workflows
<2 min
Setup time — brew install and go
100%
Local — zero cloud dependencies
Security & Privacy

Your Mac, your permissions,
your code

Monet isn't a cloud IDE or a sandboxed dev environment. It runs directly on your Mac with the same filesystem access you already have. No containers, no remote execution, no runaway agents deleting your codebase. Every Claude Code session is a real local process you can see in Activity Monitor.

100% local

Monet runs as a daemon on your Mac — not in a container, not in the cloud, not sandboxed. Same filesystem, same permissions, same tools you already use. Nothing is hosted or proxied.

No telemetry

Zero analytics, zero tracking. The only network call Monet makes is a lightweight license check — no usage data, no code, nothing else leaves your machine.

Your network only

Connect over your LAN or Tailscale. Token-based auth on every connection. No cloud relay, no port forwarding required.

Questions you're probably asking

Is my code safe?

Yes. Monet runs entirely on your Mac as a local daemon — not in the cloud, not in a container. Your code, your sessions, and your data stay on your machine. The only outbound call is a lightweight license check. No code, no usage data, no telemetry. Connect from other devices over your own local network or Tailscale.

Do I need a separate API subscription?

No. You bring your own Anthropic API key and Claude Code installation — Monet doesn't proxy API calls or add usage fees. You pay Anthropic directly for what you use, and you pay Monet once a year for the client. No double-billing, no markups, no surprise charges.

Do I need to change my workflow?

Not at all. Monet manages standard Claude Code processes under the hood. You can still use Claude Code directly in your terminal whenever you want. Monet just gives you a better way to launch, monitor, and interact with dozens of sessions at once — and do it from your phone.

What happens when the beta ends?

Your license stays valid for the full year you purchased. When beta ends, the price may increase for new customers, but your rate is locked in. And since it's a one-time annual license — not a subscription — you're never auto-billed. When your year is up, you choose whether to renew.

What if it's not for me?

7-day money-back guarantee, no questions asked. You'll know within a day whether Monet fits your workflow. If it doesn't, email us and we'll refund you immediately.

Ready to run agents
at scale?

Download it, run up to 3 concurrent agents, buy a license when you're ready.