Skip to content

Claude, already inside MONO.

Anthropic's models for writing, code and long files — no key of your own.

When a task needs careful writing, a long contract read end to end or real code, MONO can send it to Claude. You can also pick Claude by hand in the conversation. Either way, Claude works with MONO's tools on your box — your files, your board, your documents — and never sees a password or an API key.

What is MONO?

MONO is a desktop app for Windows, Mac and Linux that puts a team's conversations, people, files and data, board and browser in one window, with an AI agent working inside. Every account gets its own server — its box — where the files, databases, vault and agent live. MONO picks the best AI model for each task (Claude, GPT, Gemini and others), and it never sends, charges or publishes without a person's approval.

Who is it for?

For teams — agencies, sales, operations and professional firms — that want to work in the same files with AI inside, instead of each person alone in a chat. It also works for one person.

What MONO does with it

Long documents, careful writing and code — MONO sends it the heavy work.

  • Writes and rewrites proposals, contracts and emails in your tone
  • Reads long documents and answers with the exact clause
  • Writes and reviews code for the apps MONO builds
Never without you
  • See your keys or passwords
  • Send, publish or charge without your approval
Sending, posting and paying wait for your approval
Ask MONO
  1. 01Use Claude to rewrite this proposal so it sounds less stiff
  2. 02Read the whole contract and tell me what changes on renewal
  3. 03Have Claude review the code of my booking app
How it connects

Three steps, then just ask.

  1. 01

    Nothing to connect

    MONO already has the model behind it, with MONO's key — never yours.

  2. 02

    Ask as usual

    MONO picks the model per task, or you choose it in the conversation.

  3. 03

    Backup ready

    If a provider fails, MONO moves the task to another model.

MONO holds the account with the model provider: you don't open one, and no key of yours exists to leak. The model receives the text it needs for the task and the tools MONO allows; the credentials stay sealed in your box's vault.

  • Access sealed in your box's vault
  • Only the permissions this connection needs
  • Sending, posting and paying wait for your approval

How security works

Questions

Do I need an Anthropic account or API key?

No. MONO already has the model behind it; you just use it from the conversation.

Can I force MONO to always use Claude?

Yes: pick it in the model selector of the conversation. If you don't, MONO chooses per task.

Connect it in MONO.

Download the app, open your box and ask. MONO tells you what it needs.