Inference API terms

Last updated: November 4, 2026

These terms govern use of the Scannit Inference API, operated by Scannit Inc.

1. Access. If you use the API through a router such as OpenRouter, you also agree to that router's terms, and the router handles your account and billing. Direct access requires a separate written agreement with us.

2. Acceptable use. You must follow our Acceptable Use Policy and the license of the model you use. We may rate-limit or suspend access to protect the service or to comply with the law.

3. Your content. You keep your rights in your prompts. As between you and us, you own the outputs, to the extent the law and the model license allow. We process your content only to serve your request, as described in our Inference API data policy.

4. Confidential computing. We run the service in Intel TDX confidential virtual machines with NVIDIA GPU confidential computing. These protections reduce security risk but do not remove it entirely.

5. Availability. The service is provided "as is" and "as available", without a service-level commitment, unless a separate written agreement says otherwise. Models may be added, changed or retired.

6. Output accuracy. Model outputs can be wrong, incomplete or offensive. You are responsible for how you use them.

7. Liability. To the fullest extent the law allows:

• we are not liable for indirect, incidental, special, consequential or exemplary damages, including lost profits or data;
• our total liability is limited to the amount you paid us for the API in the 3 months before the claim.

8. Changes. We may update these terms by posting a new version here. Continued use means you accept the new version.

9. Law. These terms are governed by the laws of the State of Delaware, and the state and federal courts in Delaware have exclusive jurisdiction.

Contact: privacy@scannit.io. To report abuse or a security issue: security@scannit.io.