Three steps: get a key, point an OpenAI SDK at Huzz, send a request.

​
1. Get an API key

Sign up at huzz.ai/register, then request an API key from the Huzz team at support@huzz.ai — during early access, keys are issued by the team, with self-serve key management on the way. The raw key is shared once when it is issued — copy it and store it as an environment variable: 
export HUZZ_API_KEY="hz_..."

​
2. Point any OpenAI SDK at Huzz

Huzz speaks the OpenAI wire format, so the official OpenAI SDKs work unchanged. Set the base URL to https://api.huzz.ai/v1 and pass your Huzz key: 
import OpenAI from "openai";

const huzz = new OpenAI({
  apiKey: process.env.HUZZ_API_KEY,
  baseURL: "https://api.huzz.ai/v1",
});
No SDK? Plain HTTP works everywhere — every request just needs the Authorization: Bearer $HUZZ_API_KEY header.

​
3. Make your first chat call

curl https://api.huzz.ai/v1/chat/completions \
  -H "Authorization: Bearer $HUZZ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Say hello from Huzz!"}]
  }'
Swap "gpt-4o" for any model id in the catalog — Claude, Gemini, Grok, DeepSeek, Qwen, and more — without touching the rest of your code.

​
Next steps