Free to fork.
Metered when it runs.
Per-second compute billing, org invoices, hibernation by default, and an Apache-2.0 self-host path when you want your own machines.
Free
For prototypes, eval harnesses, and local-first teams trying PandaStack.
- $5.40 of usage credit / month (~77 hours on a 1 vCPU / 1 GiB template)
- Up to 2 vCPU / 2 GiB per sandbox
- 5 concurrent sandboxes
- 1-hour max TTL
- Hibernation included
- 1 persistent volume · 1 GiB
- Public templates
- Community Discord
Pro
For teams shipping agent features to production users.
- $20 of usage credit included every month
- Up to 8 vCPU / 16 GiB per sandbox
- 50 concurrent sandboxes
- Up to 7-day TTL
- Private templates
- 10 persistent volumes · 10 GiB each (100 GiB total)
- Fork-tree API, WS exec, MCP server
- 5 org seats
- Email support
Team
For organizations with shared billing, seats, and operational visibility.
- $99 of usage credit included every month
- Up to 16 vCPU / 64 GiB per sandbox
- 500 concurrent sandboxes
- Up to 30-day TTL
- Unlimited org seats
- 50 persistent volumes · 50 GiB each (2.5 TiB total)
- Per-seat usage attribution
- Org billing portal
- Slack support
Enterprise
For regulated, high-scale, or dedicated-infra deployments.
- Everything in Team
- BYO-VPC / dedicated single-tenant
- Unlimited concurrency + custom CPU/RAM caps
- Unlimited persistent volumes
- SAML SSO + SCIM + OIDC
- Audit-log streaming
- 99.95% SLA
- On-call
All meters charged per-second · billed monthly · CPU/RAM comes from the selected template · Pro/Team base fee converts 1:1 to usage credit · Free includes $5.40 of usage credit / month.
Billing without surprises.
The control plane records usage at sandbox stop, then submits idempotent Stripe meter events every 60 seconds.
How does per-second metering work? +
We meter active compute every second at $0.000015/vCPU/s + $0.0000045/GiB/s of RAM and submit idempotent Stripe Meter Events from the control plane after each sandbox stops. Your monthly base fee on Pro/Team converts 1:1 to usage credit — if you spend less than your base, you don't pay overage; if you spend more, the overage bills at the per-second rate.
How is a sandbox-hour computed? +
A sandbox-hour is wall-clock runtime for one active sandbox. Cost is template size × runtime: vCPU-hours at $0.054 plus GiB-hours of RAM at $0.0162, prorated per second. For example, one 1 vCPU / 1 GiB template costs $0.0702 per hour.
What counts as a sandbox? +
Each created microVM counts as one sandbox while it is active. Forks, restored snapshots, and template launches are each metered as their own sandbox from create to stop/delete.
Can I choose CPU/RAM when launching? +
No. CPU and RAM are baked into each template. Pick a template with the size you need, or bake your own template with a different CPU/memory spec.
Do hibernated or paused sandboxes bill compute? +
No. Hibernation writes state to disk and stops active vCPU/RAM metering. Storage for hibernated state is currently included at no extra charge.
How is volume storage billed? +
Every plan includes its persistent-volume quota at no extra charge (Free 1 GiB, Pro 100 GiB, Team 2.5 TiB). Provisioned storage beyond your plan's quota bills at $0.15 per GiB-month, metered continuously and drawn from the same monthly usage credit as compute. Managed database storage counts against the same meter — there is no separate database storage rate.
What's the $5.40 free credit? +
It is monthly usage credit, equal to about 77 hours on a 1 vCPU / 1 GiB template at current rates. When you exhaust it, Free workspaces pause new sandbox creation until the next month or until you link a card. Running sandboxes are never surprise-billed.
Is there an annual discount? +
Yes — annual prepay on Pro and Team saves about 17% (two months free).
Can I self-host instead? +
Yes. The core is Apache-2.0. You bring machines, networking, observability, and any cloud costs.
Can I cancel anytime? +
Yes. Monthly plans can cancel at the billing portal. Existing usage for the current period is invoiced normally.