GitHub Docs

Provider: Groq

Groq runs open-source models on custom LPU (Language Processing Unit) hardware, delivering extremely low latency inference – often 10-20x faster than cloud GPU providers.

Get an API Key

  1. Go to console.groq.com
  2. Create an account or sign in
  3. Navigate to API Keys
  4. Create a new key and copy it

Groq offers a free tier with rate limits.

Configure VibeCody

Option 1: Environment variable (recommended)

export GROQ_API_KEY="gsk_..."
vibecli --provider groq

Option 2: Config file (~/.vibecli/config.toml)

[groq]
enabled = true
api_key = "gsk_..."
model = "llama-3.3-70b-versatile"

Model Selection

Model Strengths Best for
openai/gpt-oss-120b Strong general coding Daily coding tasks
openai/gpt-oss-20b Ultra-fast responses Quick completions, simple edits
qwen/qwen3.6-27b Good balance Longer code analysis
minimaxai/minimax-m2.7 Agentic / tool use Multi-step tasks

Default: openai/gpt-oss-120b

The Llama 3.x models (llama-3.3-70b-versatile, llama-3.1-8b-instant) were deprecated on 2026-06-17 — Groq points at openai/gpt-oss-120b and openai/gpt-oss-20b as replacements. mixtral-8x7b-32768 is no longer served.

Override from the CLI:

vibecli --provider groq --model llama-3.1-8b-instant

Pricing

Groq offers a generous free tier with rate limits. Paid plans remove rate limits and add priority access.

Best For

  • Ultra-fast iteration – responses arrive in under a second
  • Interactive coding sessions – low latency makes back-and-forth feel instant
  • Running open-source models – access Llama, Mixtral without hosting them yourself

Verify Connection

vibecli --provider groq -c "Write a Go function to reverse a linked list"

Troubleshooting

Rate limited on free tier

Error: 429 Too Many Requests
  • Free tier has per-minute and per-day token limits
  • Wait 60 seconds and retry, or upgrade to a paid plan

Model not available