What fair use means

The LLM Prover API is designed for benchmarking, evaluation, and monitoring workloads. Fair use means using the API in ways consistent with that purpose.

What is not permitted

Polling GET /jobs/{job_id} faster than once every 2 seconds
Automated bulk submission that exceeds your tier’s daily request limit
Using the API to proxy or resell LLM access to third parties
Submitting prompts designed to extract training data or circumvent model safety measures
Automated account creation to bypass tier limits

What triggers suspension

Repeated rate limit violations, sustained abuse patterns, or violations of the Acceptable Use Policy may result in temporary or permanent suspension of API access.

Suspension affects API keys only – dashboard access is not affected unless the underlying account is suspended.

Staying within limits

Respect the 2-second polling interval for job status
Implement exponential backoff on 429 responses – see Rate limits
Use scheduled benchmarks instead of polling-based monitoring where possible – benchmarks run on a server-side schedule and do not consume your rate limit budget
If you need higher limits, upgrade your plan or contact support

Contact

If you believe your account has been incorrectly suspended, or if you have a legitimate use case that requires higher limits, contact support from the dashboard.