GroqCloud
Very fast hosted inference for open models, with free account-level limits.
- Requests / day
- 14.4K
- Requests / min
- 30
- Tokens / min
- 6K
GroqCloud scores 5.3/10 for free value in LLM. The main constraint: Free plan limits are shown per organization/model in Groq Console; measured as RPM/RPD/TPM/TPD.
Free value score
5.3 / 10Show the math
Key metrics
Models
16 free · 19 totalLimits may be stale — last re-verified outside this section's freshness window.
| Model | Family | Context | Limits | AA | Access |
|---|---|---|---|---|---|
| openai/gpt-oss-120b | GPT | 131.1K | 30 rpm · 8K tpm · 1K rpd · 200K tpd | 23.8 | free |
| openai/gpt-oss-20b | GPT | 131.1K | 30 rpm · 8K tpm · 1K rpd · 200K tpd | 14.9 | free |
| meta-llama/llama-4-scout-17b-16e-instruct | Llama | 131.1K | 30 rpm · 30K tpm · 1K rpd · 500K tpd | 10 | free |
| allam-2-7b | ALLaM | — | — | — | free |
| canopylabs/orpheus-arabic-saudi | Orpheus (Canopy Labs) | 4.1K | — | — | free |
| canopylabs/orpheus-v1-english | Orpheus (Canopy Labs) | 4.1K | — | — | free |
| groq/compound | Groq (system) | 131.1K | — | — | free |
| groq/compound-mini | Groq (system) | 131.1K | — | — | free |
| llama-3.1-8b-instant | Llama | 131.1K | 30 rpm · 6K tpm · 14.4K rpd · 500K tpd | — | free |
| llama-3.3-70b-versatile | Llama | 131.1K | 30 rpm · 12K tpm · 1K rpd · 100K tpd | — | free |
| meta-llama/llama-prompt-guard-2-22m | Llama | 512 | — | — | free |
| meta-llama/llama-prompt-guard-2-86m | Llama | 512 | — | — | free |
| openai/gpt-oss-safeguard-20b | GPT | 131.1K | 30 rpm · 8K tpm · 1K rpd · 200K tpd | — | free |
| qwen/qwen3-32b | Qwen | 131.1K | 60 rpm · 6K tpm · 1K rpd · 500K tpd | — | free |
| whisper-large-v3 | Whisper | — | — | — | free |
| whisper-large-v3-turbo | Whisper | — | — | — | free |
| Minimax M2.5 | MiniMax | — | — | — | paid |
| Qwen3-VL 32B | Qwen | — | — | — | paid |
What you get
Free plan exists
Units: RPM / RPD / TPM / TPD
Limits per organization/model
Very low latency inference
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Sources
-
"llama-3.1-8b-instant | 30 | 14.4K | 6K | 500K | - | - | ... llama-3.3-70b-versatile | 30 | 1K | 12K | 100K ... openai/gpt-oss-120b | 30 | 1K | 8K | 200K ... qwen/qwen3-32b | 60 | 1K | 6K | 500K ... whisper-large-v3 | 20 | 2K | - | - | 7.2K | 28.8K"
https://console.groq.com/docs/rate-limits -
"Groq free tier. Every model, no credit card. Rate-limited to 30 RPM, 6,000 tokens per minute, and 14,400 requests per day."
https://www.cloudzero.com/blog/groq-pricing/ -
"openai/gpt-oss-120b ... 30 1,000 8,000 200,000 ... moonshotai/kimi-k2-instruct 60 1,000 10,000 300,000 ... The Free Tier Rate Limits (March 2026) ... pulled directly from Groq's official documentation."
https://www.grizzlypeaksoftware.com/articles/p/groq-api-free-tier-limits-in-2026-what-you-actually-get-uwysd6mb -
"gpt-oss-120b (high) ... Intelligence: 33 Artificial Analysis Intelligence Index"
https://artificialanalysis.ai/models/gpt-oss-120b -
"Groq ... today announced $750 million in new financing at a post-money valuation of $6.9 billion"
https://techcrunch.com/2025/09/17/nvidia-ai-chip-challenger-groq-raises-even-more-than-expected-hits-6-9b-valuation/
Alternatives
Same category · similar quota
Scaleway's EU-sovereign serverless Generative APIs give every new customer 1,000,000 free tokens (plus 60 free audio-transcription minutes) across open models like Qwen, Llama, Mistral and gpt-oss.
