Pricing
Pay only for the GPU-minutes you use
No idle license fees. No cluster to amortize. Start with a guided demo, scale to multi-GPU on demand, and cap your spend with hard budget limits.
Demo
See it on your own geometry, with our team.
- Live 30-minute guided walkthrough
- We run your geometry with you
- Trial GPU credits to try it yourself
- Browser 3D visualization of results
- Solver & workflow Q&A
- No credit card
Pro
Usage-based — pay only while you're solving.
- Pay-as-you-go, billed by the minute
- Rate scales with GPU tier (T4 → A100 → H100)
- Multi-GPU solving
- Unlimited mesh size (per GPU tier)
- Parameter sweeps & mesh studies
- Budget caps + spend alerts
- 90-day retention + exports
- Up to 10 team members
- Email support
Enterprise
Security, scale, and support for your org.
- Multi-node H100 clusters
- SSO / SAML & SCIM
- VPC or on-prem deployment
- Custom budgets & invoicing
- Audit logs & access reviews
- Signed DPA
- Dedicated solutions engineer
- Priority SLA support
The Pro rate shown is a starting point — the exact GPU-minute rate depends on the GPU tier you choose. All plans include hard budget caps and spend alerts.
Compare
Plan comparison
The details that matter when you scale from a single study to a team's worth of cases.
| Feature | Demo | Pro | Enterprise |
|---|---|---|---|
| GPU types | T4 (guided) | T4 / A100 / H100 | All + reserved |
| Multi-GPU | — | Up to 8× | Multi-node |
| Concurrent runs | Guided | 10 | Custom |
| Budget caps | — | Yes | Org-wide |
| SSO / SAML | — | — | Yes |
| Deployment | Shared cloud | Shared cloud | VPC / on-prem |
| Support | Demo call | Dedicated + SLA |
FAQ
Pricing questions
A GPU-minute is one minute of wall-clock time on one GPU while your case is actively solving (billed by the minute). The per-minute rate depends on the GPU tier — T4 is cheapest, H100 the fastest. Multi-GPU runs bill per device, and idle, queued, and post-processing time is not metered.
See it on your own case
Book a 30-minute guided demo, then jump straight into pay-as-you-go Pro.