← Back to LighterHub

Terms of Service

Last updated: 2026-05-27

These Terms govern your use of the LighterHub API and website (collectively, the "Service"), operated as LighterHub ("LighterHub", "we", "us"). By using the Service or arranging paid access, benchmark work, or reserved capacity with LighterHub, you agree to these Terms.

1. The Service

LighterHub provides planned access to open-weight large language models hosted on GPU hardware. The Service may expose an OpenAI-compatible HTTP API after workload review, benchmark, capacity planning, and operational fit checks. Available models, context windows, capacity terms, and prices may vary by written quote.

2. Access & API keys

You receive an API key only after LighterHub approves or quotes access for a workload. You are responsible for keeping your key confidential. Anyone in possession of the key can use the access associated with it. You may rotate or revoke your key by contacting founder [at] lighterhub.app.

3. Capacity quotes & billing

4. Refunds

See our Refund Policy for details. In short: paid workload reviews, benchmark work, evaluation access, and quoted capacity purchases are final and non-refundable once work begins or access is provisioned, except for duplicate charges, billing errors, failure by LighterHub to provide the written paid scope, or refunds required by law.

5. Acceptable use

You may not use the Service to:

We may suspend, throttle, or terminate any API access violating these rules. Paid access tied to terminated usage is not refundable except where required by law.

6. Service availability

The Service may run on dedicated GPU capacity that can be started, stopped, reserved, or reconfigured based on customer demand. We do not currently offer a formal SLA for shared access. We make commercially reasonable efforts to keep active routes available, but planned maintenance, hardware failures, upstream provider issues, or GPU availability may cause temporary unavailability. Availability expectations for reserved or dedicated capacity are handled in the applicable customer terms.

7. Privacy & data handling

Request and response payloads are not logged by default. We log operational metadata such as timestamps, token counts, request IDs, status, latency, model or endpoint, and your API key prefix for billing, support, abuse detection, and reliability. Temporary debug logging, if ever enabled for your requests, is handled under the notice and time limits in the Privacy Policy.

8. Disclaimers

The Service is provided "as is" without warranties of any kind. Model outputs may be inaccurate, biased, or unsafe; you are responsible for checking outputs before relying on them. We are not liable for any decisions or actions taken based on model output.

9. Limitation of liability

To the maximum extent permitted by law, LighterHub's aggregate liability for any claim arising from the Service is limited to the amount you paid for the affected quoted access, benchmark work, or capacity window in the 30 days preceding the claim.

10. Changes

We may update these Terms. Material changes will be announced on the website and via email to active customers at least 7 days before taking effect. Continued use of the Service after the effective date constitutes acceptance.

11. Contact

Questions about these Terms: founder [at] lighterhub.app