Pricing
Open platform.
BYOK. No surprises.
Run the whole substrate on your own inference keys. We don't middleman the model call, don't resell tokens, don't gate the editor. We charge for capacity, audit, and procurement — the parts that actually take operations work to keep up.
Solo
Free$0/mo
Personal projects on your own keys.
For one operator who wants the agent on their own keys.
- BYOK inference (no managed keys)
- 1 active edit at a time
- 10-min Sandbox max
- Community support
Studio
Recommended$29/mo
Pro work with managed Sandbox capacity and concurrency.
For solo operators who ship every week and want headroom.
- Everything in Solo
- 5 concurrent vibe-edits
- 45-min Sandbox max
- Marketplace publishing
- Email support
Team
Paid$99/mo per seat
Shared workspaces with audit trails and admin controls.
For studios of 2–20 with shared projects + audit needs.
- Everything in Studio
- Shared projects + handoff queue
- Reserved Sandbox capacity
- SSO + audit log export
- Priority support · 4-hour business-hours response
All paid tiers · 14-day refund window · Cancel anytime
Enterprise
Procurement-friendly.
Annual contracts shaped around seat count, Sandbox capacity, support tier, and data residency. Net 30 or Net 90 billing. Purchase orders accepted. Quote in two business days.
- Everything in Team
- Net 30 / Net 90 invoicing
- Custom seat counts + cross-team workspaces
- 99.95% uptime SLA · 1-hour P0 response
- Dedicated Slack/Teams support channel
- SSO (SAML), audit log streaming, SCIM
- Custom domains + reserved Sandbox capacity
- Annual billing · purchase orders accepted
Compare
Side-by-side.
| Capability | Solo $0 /mo | Studio $29 /mo | Team $99 /mo per seat | Enterprise Custom annual |
|---|---|---|---|---|
| Active concurrent edits | 1 | 5 | 15 | Custom |
| Sandbox max duration | 10 min | 45 min | 2 hours | Custom |
| Reserved Sandbox capacity | ||||
| Marketplace publishing | ||||
| Shared workspaces + handoff queue | ||||
| Audit log export | Streaming | |||
| SSO (SAML) | ||||
| SCIM provisioning | ||||
| Custom domains | 1 | 5 | Unlimited | |
| Net 30 / Net 90 invoicing | ||||
| Uptime SLA | Best-effort | 99.5% | 99.9% | 99.95% |
| Support response (P0) | Community | Email · 24h | Email · 4h biz | Slack · 1h, 24/7 |
FAQ
Questions before checkout.
What does BYOK actually mean here?
Bring Your Own Keys. You add your Anthropic / OpenAI / OpenRouter / AI Gateway key in account settings. The platform's agent calls run against your key, charge your account, and land in your provider's audit log. We never proxy inference, never aggregate, never resell.
Can I switch tiers mid-cycle?
Yes. Upgrades prorate immediately and unlock features on the next API call. Downgrades take effect at the end of the current billing period — you keep what you paid for through the end of the cycle.
Do you offer refunds?
First 14 days on Studio and Team are no-questions refunded if you ask. After that we issue prorated refunds for billing errors and service-credit-eligible SLA misses (see /sla). We don't claw back usage you already had.
How does the Team per-seat number work?
$99 covers one operator seat per month. Add or remove seats from /billing — additions prorate, removals take effect next cycle. Cross-team workspaces, SCIM, and SSO move you onto Enterprise where seat handling is part of the contract.
What's included in Enterprise pricing?
Annual contract sized to your seat count, Sandbox capacity, support tier, data residency, and any custom SLA. Net 30 / Net 90 invoicing, purchase orders, and SOC-2-aligned security review are standard. Quote within 2 business days of /enterprise form submission.
Why is GAD priced this way? You're cheap on the lower tiers.
Inference is the cost driver, and you're already paying for it on your own keys. We don't double-charge for the model call. We make money on Marketplace publishing, Team's reserved Sandbox capacity, and Enterprise contracts — not by gating the editor.