Terms of Service.
1. Who we are
These terms are between you and GGEZ Enterprise, osoba rizikového kapitálu, s.r.o., Proskovická 689/100, Výškovice, 700 30 Ostrava, Czech Republic, IČO 23173602, DIČ CZ23173602, registered at the Regional Court in Ostrava under file C 99342 ("we"). Contact: [email protected].
2. The service
We operate inference endpoints that serve open-weight language models. The models, their
context limits and their prices are declared at our /v1/models endpoint, which is
the authoritative list — anything else describing the service is a summary. Requests typically
reach us through OpenRouter, which routes traffic and handles end-user billing.
3. Availability
The service is provided on a best-effort basis. We aim to keep endpoints reachable and can shed load under abuse or overload, but we do not promise an uptime figure unless we have agreed one with you in writing. Capacity is finite and shared; identical requests may be routed elsewhere when we are full.
4. Acceptable use
You agree not to use the service to:
- break any law, or infringe anyone's rights, including intellectual-property rights;
- generate or distribute sexual content involving minors, or material that is illegal in the Czech Republic or in your jurisdiction;
- attack, overload, probe or reverse-engineer our infrastructure, or attempt to bypass rate limits or safety controls;
- present the output as professional legal, medical, tax or financial advice to a third party without independent human review;
- resell raw access to our endpoints as a compute service without a written agreement with us.
Selling a product that embeds our API is fine. Reselling the endpoints themselves as your own GPU capacity is what needs an agreement first — ask, and we will usually say yes.
5. Your responsibilities
You are responsible for the content you send, for having the right to send it, and for what you do with the output. Models are distributed under their own licences (for example Apache-2.0 for the Qwen family); those licences bind you as well, and you must respect any use restrictions the model's authors set.
6. Pricing and billing
Prices are per token and are declared in the model listing. Where you reach us through OpenRouter, OpenRouter bills you and settles with us; where you call us directly, we invoice you. We may change prices for future usage — a change never applies retrospectively to usage already served.
7. No warranty
Large language models produce probabilistic output that can be wrong, incomplete or offensive. The service is provided "as is", without warranties of any kind, including fitness for a particular purpose, accuracy, or uninterrupted availability. Verify anything that matters.
8. Limitation of liability
To the maximum extent permitted by law, we are not liable for indirect or consequential loss, lost profits, lost data or business interruption. Our total liability arising out of the service is limited to the amount you paid us for it in the three months before the event giving rise to the claim. Nothing here limits liability that Czech law does not allow to be limited.
9. Data protection
We do not retain prompts or completions. What we process, where it happens and your rights are set out in our Privacy Policy, which forms part of these terms.
10. Suspension and termination
We may suspend or terminate access if you breach section 4, if we are legally required to, or if your traffic threatens the service for others. You may stop using the service at any time.
11. Changes
We may update these terms; material changes take effect when we publish the new version with a new effective date. Continuing to use the service after that means you accept them.
12. Governing law
These terms are governed by the laws of the Czech Republic. Disputes are subject to the jurisdiction of the Czech courts, with the courts in Ostrava as the court of first instance.