Groq Free AI Models - 3 Models, 30 RPM / 1,000 RPD
Groq Free AI Models - 3 Models, 30 RPM / 1,000 RPD
GroqCloud’s free plan is available for text models with no credit card required, and is rate-limited by requests and tokens.
Free-plan limits (official November 2026 data): 30 RPM / 1,000 RPD / 8K TPM / 200K TPD, calculated per organization, not per API key. Access from mainland China requires a proxy.
Model List (3)
| Model | Context | Max output | Description |
|---|---|---|---|
| GPT-OSS 120B | 131K | 65K | OpenAI’s 120B open-source reasoning model, ~500 tokens/s |
| GPT-OSS 20B | 131K | 65K | 20B lightweight version, ~1,000 tokens/s |
| Qwen 3.8 27B | 131K | 16K | Alibaba’s 27B multimodal model with image input, reasoning, and tool calling |
The speech-to-text model
whisper-large-v3has a free-plan limit of 20 RPM / 2,000 RPD; it is not listed separately here.
Usage Example
curl https://api.groq.com/openai/v1/chat/completions \
-H "Authorization: Bearer $GROQ_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-oss-120b",
"messages": [{"role": "user", "content": "Hello!"}]
}'Free Plan Limits
- Rate limits: 30 RPM / 1,000 RPD / 8K TPM / 200K TPD, whichever is hit first
- Concurrency: low on the free plan; high concurrency returns 429
- Data: free-plan data retention follows Groq’s official policy; do not submit sensitive information
- Network:
api.groq.comis not directly reachable from mainland China
Related Links
Last updated on
· Source: Groq official documentation and pricing pages. Quotas and prices may change at any time.