Hi NVIDIA team,
I’m unable to call deepseek-ai/deepseek-v4-flash-0731 via the NIM API at
https://integrate.api.nvidia.com/v1.
- GET /v1/models returns 200 and lists the model.
- POST /v1/chat/completions with this model returns 504 after ~300s, 0 bytes.
- Other third-party models return 404 “Not found for account ‘…’”.
- First-party nvidia/* models work instantly with the same key.
- The model works fine in the build.nvidia.com playground.
Account email: yousef.hamza.deeb@gmail.com
API Key prefix: nvapi-g7PCfuFG7mlG5ARcWYh***********************************
Model: deepseek-ai/deepseek-v4-flash-0731
Could you check whether my personal organization has the “Public API Endpoints”
permission enabled, or enable it for this model? Thanks.