Terms of Service

These Terms govern the Open Scale inference infrastructure (the “Service”). Open Scale operates the Service as a model-inference provider to platforms such as OpenRouter. We do not offer user accounts, issue API keys to end users, or bill end users directly. End users interact with the Service through the platform they use, and that platform’s own terms and policies govern the end-user relationship.

1. The service

We provide programmatic access to open-weight LLMs through an OpenAI-compatible API. Platforms send a request; we route it to suitable compute, run inference, and return a response with usage information. We are an inference provider. We do not own or develop the underlying open models, which remain subject to their respective licenses.

2. Relationship and eligibility

The Service is provided to the platforms and business partners that route traffic to us. If you access the Service through a platform such as OpenRouter, your use is also subject to that platform’s terms, and the platform is responsible for the accounts, credentials, and billing of its users. We do not create or manage end-user accounts or credentials.

3. Capacity & availability

Inference capacity is finite. When the compute capacity for a model is fully utilized, we return a 429 (Too Many Requests) response rather than queueing the request, with a Retry-After hint. This is expected behavior, not an outage. We work to maintain high availability and provide a target uptime.

4. Pricing

• Pricing is per token (input, output, and cached reads) and is published per model.
• Cached input tokens may be billed at a reduced rate; usage is reported in the response.
• We may change pricing; material changes are reflected before they take effect.
• End users are billed by the platform they use, not by Open Scale. We do not process end-user payments.

5. Acceptable use

Traffic sent to the Service must not:

• Generate content that is illegal, harmful, or that infringes others’ rights.
• Reverse-engineer, probe, or abuse the API or the underlying models.
• Attempt to disrupt, overload, or gain unauthorized access to the infrastructure.
• Misrepresent the Service or its terms when directing traffic to us.

6. Data & content

Rights to content submitted through the Service and to outputs generated by the Service belong to the platform or its users, subject to the underlying model licenses. We process that content only to provide the requested inference and do not retain it in persistent storage. See our Privacy Policy for details.

7. Output quality & limitations

Models are probabilistic. Outputs may be inaccurate, incomplete, or inappropriate. We provide the Service “as is” and the platform and its users are responsible for reviewing and validating outputs before relying on them, especially for high-stakes uses.

8. Suspension

We may suspend or terminate access for violations of these Terms, abuse, or where required by law, with or without notice.

9. Disclaimer of warranties

The Service is provided “as is” without warranties of any kind, express or implied, including fitness for a particular purpose and non-infringement.

10. Limitation of liability

To the maximum extent permitted by law, our total liability arising out of or relating to these Terms is limited to the amounts paid to us for the Service in the twelve months before the claim. We are not liable for indirect, incidental, or consequential damages.

11. Changes & governing law

We may update these Terms; continued use after changes constitutes acceptance. These Terms are governed by the laws of the State of Wyoming, United States.

12. Contact

For questions about these Terms, contact us at legal@openscalecloud.com.