Contact
One inbox, read by the person who runs the service.
API errors, balance questions, refunds, key issues. Include your account email and, for API problems, the request id from the error response.
hello@inferenceapis.comWant a model we do not serve yet? Tell us the exact id and what you are building. Open-weight models that run on standard GPUs are usually a few days.
Request a modelA provider changed a limit, retired a model, or an entry in the error reference is out of date. Send the URL and what changed; we re-verify and update the entry.
Report a changeFound a vulnerability? Email with "Security" in the subject. We acknowledge within two business days and do not pursue good-faith researchers.
Report a vulnerabilityTypical response time is same business day. Service availability is on the status page; recent changes are in the changelog.