ModelHubby

← All guides · provider record →

Groq — free tier setup

What you get free

A permanent free tier on Groq's ultra-fast LPU inference, with per-model rate limits applied at the organization level. Per the catalog: "llama-3.3-70b-versatile: 30 RPM, 1K RPD, 12K TPM, 100K TPD; llama-3.1-8b-instant: 30 RPM, 14.4K RPD, 6K TPM, 500K TPD; openai/gpt-oss-120b: 30 RPM, 1K RPD, 8K TPM, 200K TPD (varies by model). 'Rate limits apply at the organization level, not individual users.'" The gotcha: Groq cut free-tier limits in 2026 — most models are now 1,000 requests/day (down from 14,400), so daily caps arrive much sooner than older guides suggest.

Friction check
Credit card no
Phone verification no
Org/membership none
Est. time to first key ~5m (3 steps)
Trains on your prompts unknown

Get a key

  1. Go to console.groq.com and sign up (email or OAuth) — no card required.
  2. Open the API Keys page at console.groq.com/keys. The official quickstart says to "visit the API Keys page to create an API Key."
  3. Create a new key (exact button label unconfirmed — verify on first signup), name it, and copy the key immediately. It is shown once.

First request

curl against the OpenAI-compatible base URL from the catalog. Expected success shape: a JSON chat.completion object with choices[0].message.content.

export GROQ_API_KEY="YOUR_KEY"
curl https://api.groq.com/openai/v1/chat/completions \
  -H "Authorization: Bearer $GROQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.1-8b-instant",
    "messages": [{"role": "user", "content": "ping"}],
    "max_tokens": 16
  }'

Limits that will bite you

  • Limits are per model, at the organization level — a second key in the same org does not add quota.
  • llama-3.3-70b-versatile: 30 RPM, 1,000 RPD, 12K TPM, 100K TPD — the 100K TPD ceiling is the tightest for long-context work.
  • llama-3.1-8b-instant: 30 RPM, 14,400 RPD, 6K TPM, 500K TPD — highest RPD on the tier; the 6K TPM cap throttles bursts.
  • openai/gpt-oss-120b: 30 RPM, 1,000 RPD, 8K TPM, 200K TPD. qwen/qwen3-32b: 60 RPM, 1,000 RPD, 6K TPM, 500K TPD.
  • 2026 reduction: most models now 1,000 RPD on free tier. Cached tokens do not count toward thresholds.
  • Live model list: https://api.groq.com/openai/v1/models.

Key rotation runbook

  • Create a second key from the same API Keys page (console.groq.com/keys) (multiple-keys support unconfirmed — verify on first signup). Note: extra keys do not raise quota — limits are org-level, so real rotation requires a second organization/account.
  • Revoke: delete the key from console.groq.com/keys.
  • ModelHubby integration: export GROQ_API_KEY_2="..." — the key pool picks it up automatically; state machine cools down rate-limited keys per-key.

ModelHubby usage

pnpm mh show groq
pnpm mh probe --live --provider groq --apply