Hi,
I’m consistently getting HTTP 429 responses from the NVIDIA NIM API.
Error:
HTTP 429
{
“status”: 429,
“title”: “Too Many Requests”
}
Environment:
- Using the NVIDIA NIM API
- Model: GLM-5.2 (also interested to know if this is model-specific)
- Client: Custom AI coding agent / CLI
- Requests are sequential (not intentionally flooding the API)
- Automatic retries (3 attempts) all return 429
What I’ve checked:
- https://status.ngc.nvidia.com shows no ongoing incidents.
- My API key is valid.
- The same workflow was working previously without changes.
Questions:
- Is there currently a rate limit or quota issue affecting GLM-5.2 or NIM endpoints?
- Are there hidden per-minute or concurrent request limits that could trigger a 429 even with available credits?
- Is this likely an account-specific issue or are other users experiencing the same behavior?
- Is there a way to inspect my current rate limits or remaining quota?
Any guidance would be appreciated. Thanks!