Groq

Overview

Groq offers ultra-fast LLM inference — Llama, Mixtral, and other open models served at speeds far beyond typical GPU clouds — via an OpenAI-compatible API. Auth is a free API key (an invalid key returns 401, verified).

Quick Start

List models (replace YOUR_KEY):

curl "https://api.groq.com/openai/v1/models" \
  -H "Authorization: Bearer YOUR_KEY"

Chat completion (OpenAI-compatible):

curl "https://api.groq.com/openai/v1/chat/completions" \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.3-70b-versatile",
    "messages": [{"role": "user", "content": "Explain quantum computing in one sentence"}]
  }'

Authentication

• Register at console.groq.com → free API key.
• Send it as Authorization: Bearer YOUR_KEY (an invalid key returns 401 — verified).

Rate Limits & Notes

• Free tier includes monthly token allowances per model (check the console) — great for prototyping.
• OpenAI-compatible: switch the base URL to https://api.groq.com/openai/v1 in OpenAI SDKs.
• Model IDs change over time — fetch /models for the current list.
• Known for very low latency/token — popular for agents and streaming UIs.

FAQ

Is Groq free? Free tier with monthly token allowances; paid plans scale.

How do I get a key? Register at console.groq.com — the key is on your dashboard.

Is it OpenAI-compatible? Yes — point OpenAI SDKs at https://api.groq.com/openai/v1.