Free
€0$0£0
Everything you need to build, train, and run on your own GPU.
- Unlimited local workflows on your own GPU
- Local-first and fully offline-capable
- Author and share your own extensions
- Community support
Pricing
Free to run. Priced to grow.
The platform runs locally on your hardware, so the core is free. Paid plans add collaboration, scale, and support — never a paywall on the compute itself.
Plans
Pro
€19$19£16 / month
For individuals and small teams who collaborate.
- Everything in Free
- Shared workspaces and team libraries
- Private extension sharing
- Priority support
Enterprise
Custom
For organizations with compliance and scale requirements.
- Everything in Pro
- SSO, audit logs, and role controls
- Dedicated support and SLAs
- Deployment and onboarding help
Prices shown per seat. Local compute is always free — you only ever pay for collaboration and support.
Frequently asked questions
Does my data leave the device?
No — not by default. Computation runs locally, and any network access is a declared, consent-gated effect you approve in advance.
Do I need a special GPU?
No. Compute runs on WebGPU — the GPU already in your machine. There is no CUDA requirement and no vendor lock-in.
Can I extend the platform?
Yes. You can author blocks, kernels, viewers, and models in-app; user content is structurally identical to the built-ins.