it's free*.ai

NVIDIA NIM

The biggest free catalog: 82 open models over API.

Get API keyOfficial docs ↗

NVIDIA NIM on build.nvidia.com is the evaluation endpoint for NIM microservices: one OpenAI-shaped API in front of the widest free catalogue here. The live /v1/models list is 82 ids — DeepSeek V4 Flash, Nemotron 3 Ultra, Laguna XS, Kimi K3, MiniMax M3.

A phone number, never a card. Up to 40 requests a minute and a million tokens of context on the long models. GLM 5.2 is not on the current list. NVIDIA says it is for evaluation, not production traffic — treat the key that way.

Free models
82
Max context
1M
Free tier
40 requests per minute
Requirement
Phone verification, no card

Endpoint

Base URLhttps://integrate.api.nvidia.com/v1
Env varNVIDIA_API_KEY
Rate limitUp to 40 requests/minute

Code examples

npm install openai
import OpenAI from 'openai'

const client = new OpenAI({
  baseURL: 'https://integrate.api.nvidia.com/v1',
  apiKey: process.env.NVIDIA_API_KEY
})

const response = await client.chat.completions.create({
  model: 'deepseek-ai/deepseek-v4-flash-0731',
  messages: [{ role: 'user', content: 'Explain closures in one paragraph.' }]
})

console.log(response.choices[0].message.content)

Worth knowing

Models you get for free82