Unlimited coding on Capuchin. Any other model on your terms.
$20/month includes unlimited use of Capuchin — our own model — plus Claude, Gemini, ChatGPT on your own key, or your own local/remote model. A hard spend cap means your invoice never surprises you.
Free
Full agentic IDE on all platforms — local-first, any model, no time limit.
Pro
Unlimited coding on Capuchin. Any other model on your terms.
Pro+
3× frontier allowance, priority queue, more parallel agents.
Ultra
For power users and frontier-heavy workflows.
Team — $30/seat/mo
Everything in Pro, plus shared encrypted team index, roles, allowlists, version pinning, and centralized billing with usage by member, project, and model.
Start a teamThe cheapest token is the one you don't pay us for.
Unlimited on Capuchin
We run our own model on our own infrastructure — generous usage is cheap for us, free at the margin for you.
BYOK frontier
Use Claude, Gemini, or ChatGPT on your own key and pay your provider directly. We don't mark up your tokens.
Local & self-host
Run Capuchin or any model on your own hardware and pay near nothing per token.
Hard cap
Set a ceiling; agents stop instead of billing you — the opposite of usage-based tools whose bills jumped 10–50× for power users.
Frequently asked questions
Is Free time-limited?
No — the Free tier is built to actually be used, not as a time-gated trial.
Which models can I use?
Capuchin (ours), Claude, Gemini, ChatGPT, or your own — local or on your own server. BYOK on every tier.
What does 'unlimited' mean?
Unlimited Capuchin usage under a fair-use ceiling that only affects extreme volume. Frontier models are BYOK or at-cost.
When do the Terminal and desktop app arrive?
~1 month after the IDE launch, included on your plan at no extra cost.
Can I run with no internet?
Yes — fully air-gapped on Enterprise / Self-Host.
Is it open source?
Open core, Open VSX, MIT framework underneath. Your code is yours.