Are there limits on the number of API requests?
If an operational impact occurs, such as an overload of the entire service caused by a large number of concurrent requests, we may impose limits on the number of concurrent requests. Requests that exceed the limit will receive HTTP status 429.
For usage constraints, see the "Notes" section of "Generative AI Service" in the User Guide.
What should I do if responses are slow or not returned?
Temporary delays, timeouts, or 502 errors may occur when requests from users are concentrated. Please try again after some time. If the problem persists, please contact support with the time of occurrence, the model name, and the error details.
Related information
- User Guide "Generative AI Service"
- For pricing, see "Generative AI Service Pricing"
- For data handling, see "Data Handling and Use of External Tools in the Generative AI Service"
If this article does not solve your issue
Please contact support with the time of occurrence, the model name, and the error message (including the HTTP status code).