Hi NVIDIA team,
I'm unable to call third-party NIM models (e.g., deepseek-ai/deepseek-v4-pro, deepseek-ai/deepseek-v4-flash,minimaxai/minimax-m2.7) via https://integrate.api.nvidia.com/v1. they are tested ok this afternoon, but in the evening, they don't work.
- These models work fine in the build.nvidia.com playground with the same account
- nvidia/nemotron-3-super-120b-a12b works fine via API
- /v1/models endpoint returns 200 and lists the models
- POST /v1/chat/completions times out after 120s (returns no data)
I have already regenerated my API key from build.nvidia.com, but the issue persists.
Could you please check if my personal organization has the "Public API Endpoints" permission enabled?
Can you tell me how to correct it?
**Account email**: davidyu_29@163.com
**API Key prefix**: nvapi-ocl-KRswQ1m5O******
Thanks for your support.
Best regards
David