Critical API Latency and 404 Errors - Request to enable Public API Endpoints

Hi NVIDIA Developer Support Team,

Primary Issue: Severe Latency
My main problem is the extreme delay when using deepseek-ai/deepseek-v4.1-flash. Even for a simple prompt like “hello”, the model takes 5 to 10+ minutes to respond. This makes development completely impossible.

Secondary Issue: 404 Not Found
Additionally, I am unable to use moonshotai/kimi-k3, which returns the following error:

  • Status: HTTP 404 Not Found

  • Detail: Function '23d4f03a-b8a6-4adb-a183-7daa083a09cc': Not found for account 'cD6wQyBuLKChzWdzfrpcQY_cKT4JwAKuNamGYuCPFx8'

My Account ID: cD6wQyBuLKChzWdzfrpcQY_cKT4JwAKuNamGYuCPFx8

What I have verified:

  • API key is freshly generated and listed as ACTIVE on build.nvidia.com (org: personal, API User verified).

  • DNS/TLS to integrate.api.nvidia.com is fine.

  • GET /v1/models returns 200 with my key, so the key itself is valid.

  • The request payload follows the reference code exactly.

  • I am based in Ukraine and registered with a valid phone number.

From reading other threads in this category, it looks like my personal organization is missing the “Public API Endpoints” service/role. I suspect this missing role is not only causing the 404 error but may also be causing the severe latency on DeepSeek (perhaps the API is attempting to call an unauthorized function and timing out before falling back).

Could you please enable “Public API Endpoints” for my account/organization?

Thank you for your time and assistance.

Best regards,