// pricing
Two plans. Claude's usage bills to you, not through us.
Start with the tracker on its own. Move to Pro when you want Claude doing some of the work.
Free
$0
For one person keeping track of projects.
- Projects, task lists, tasks and to-dos
- Notes on every task
- Search across everything, light and dark themes
Pro
$15 a seat, each month
Everything in Free, and Claude at work on it.
- Claude runs in three modes: read, edit files, run commands in a sandbox
- Schedules on a cron, with retries, limits and Run now
- GitHub and upload workspaces; commits, pushes and pull requests
- Approval gates before a run and before a commit
- The encrypted vault for your keys and tokens
- MCP endpoint and scoped API tokens
- Your team: invitations, roles, more than one organization
- Monthly AI budget, cost per run, email notifications
// questions
The fine print, in plain words.
Who pays for Claude?
You do, directly. Runs use an Anthropic API key you add to the vault, so model usage appears on your Anthropic bill. qubichelm shows the cost of every run and lets you cap the month.
What's a seat?
A person in your organization. You buy enough seats to cover everyone in it and add more as the team grows. Checkout and invoices go through Stripe.
Do I need a GitHub repository?
No. Claude can work on files you upload, or simply read your projects and report. A repository is for when you want commits and pull requests.
Where does it run?
qubichelm is a hosted service at app.qubichelm.com; there's nothing to install. Command-running work happens on a separate, isolated runner. The security page has the details.
Can I stop a run that's spending too much?
Yes: cancel it from its page, give each schedule a time limit, require approval before scheduled runs start, and set a monthly cap that stops runs when it's reached.