Hello NVIDIA Support Team,
I am testing hosted inference via integrate.api.nvidia.com with an active NVIDIA Developer Program account.
Diagnostics:
• GET /v1/models → HTTP 200 (82 models returned).
• POST /v1/chat/completions → Times out with 0 bytes received across multiple models (minimal text payload tested on Kimi K3 and Nemotron 3.5 Lightning).
• Web Playground (build.nvidia.com) → Shows network/CORS failure on “Try It!”.
Could you please verify whether my organization has the “Public API Endpoints” hosted inference entitlement enabled, or if there is an active routing block on this account?
Account Details:
• Organization Name: MIKART
• Account Number: 1049736935712816
• Registered Email: mikias.mikart@gmail.com
• API Key prefix: nvapi-