Updates, new features, and always the latest AI models. No token fees.
The software license is billed yearly per GPU and covers the stable, secure, and current operation of your on-premise AI. Unlike the cloud, you pay fixed, predictable costs, independent of tokens, API calls, or user counts.
Core service: model updates
- Curated model selection (state of the art)
- Continuous model updates: new models, better performance
- Ongoing evaluation and replacement of outdated models
You always work with the latest available models, without any update or migration effort on your side.
Also included
- Security updates and new features across the entire platform
- Optional remote monitoring and support (GPU, memory, latencies, system health)
After the first year you decide whether to renew the license. Your infrastructure stays operational either way.
Not included
- Hardware (servers, GPUs): purchase or rent, see the pricing model
- Power and cooling
- Custom development and integration (adoption, billed by effort)
- On-site visits (unless explicitly agreed)
All-inclusive: Managed On-Premises
Prefer not to deal with operations at all? With Managed On-Premises you get hardware, operations, and monitoring from a single source, from €1,490 per GPU per month. Details on the pricing page.
In short
Predictable costs, no token billing, no surprises, and always current models with full cost control. What that means for your team is shown by the cost calculator.
Next steps
- Open the cost calculator and run your scenario
- Contact us for individual advice on license and operations
Sources and further information:
- How much does it cost to run AI yourself? (pricing model and building blocks)
- Can I start small and scale later? (scaling)
- Maintenance and updates (details on updates and remote maintenance)
- On-premise AI for SMEs (hardware and operations)