DOCSEARCHbyAlgolia
Submit Docs

NVIDIA API Catalog

website·Indexed 19 days ago

https://build.nvidia.com/

Documentation for the NVIDIA API Catalog, featuring NVIDIA NIM hosted, OpenAI-compatible model APIs for inference, vision, speech, retrieval, and more.

Tokens
231
Snippets
2
Records
2
Agent score
50%

What's inside NVIDIA API Catalog

  1. Authenticate and connect to NVIDIA NIM APIs

    All NVIDIA NIM APIs use a common base URL and Bearer token authentication. To use these APIs, generate an API key at https://build.nvidia.com/settings and include it in the request header as Authorization: Bearer $NVIDIA_API_KEY.
    # Base URL: https://integrate.api.nvidia.com/v1
curl https://integrate.api.nvidia.com/v1/chat/completions \
  -H "Authorization: Bearer $NVIDIA_API_KEY" \
  -d '{"model":"nvidia/llama-3.1-70b-instruct","messages":[{"role":"user","content":"Hello"}]}'

  2. Use OpenAI SDKs with NVIDIA NIM

    All NIM model endpoints hosted at https://integrate.api.nvidia.com/v1 are compatible with the OpenAI Chat Completions API. To integrate, point an existing OpenAI SDK client to the NVIDIA base URL and specify the desired NVIDIA model name.
    // Example configuration for an OpenAI-compatible client
const client = new OpenAI({
  baseURL: 'https://integrate.api.nvidia.com/v1',
  apiKey: process.env.NVIDIA_API_KEY,
});