Provider
Groq
9 Models · 13 Prices · Cheapest on 3 · Uptime 100% · updated 2026-11-03 07:16 UTC
Trust signals · SLA · model-version transparency · pricing stability · quantization-degradation · incident response · data privacy — measured by TKX probe (activating).
About
Groq is a machine-learning system that provides AI inference services, serving 14 models via platforms like Hugging Face, official repositories, and OpenRouter. It specializes in deploying and optimizing machine learning models for various applications.
AI-generated summary from public information — is this yours? Claim & edit
Models served
| Model | $ in /1M | $ out /1M | Cheapest | Router / Source | Evidence |
|---|---|---|---|---|---|
| llama-prompt-guard-2-22m | $0.030 | $0.030 | ✓ | Pricebook (LiteLLM) ↗ | Official list |
| llama-prompt-guard-2-86m | $0.040 | $0.040 | ✓ | Pricebook (LiteLLM) ↗ | Official list |
| gpt-oss-20b | $0.075 | $0.300 | — | Pricebook (LiteLLM) ↗ | Official list |
| gpt-oss-safeguard-20b | $0.075 | $0.300 | ✓ | Pricebook (LiteLLM) ↗ | Official list |
| llama-guard-3-8b | $0.200 | $0.200 | — | Pricebook (LiteLLM) ↗ | Official list |
| gpt-oss-120b | $0.150 | $0.600 | — | Pricebook (LiteLLM) ↗ | Official list |
| Llama 3.3 70B Instruct | $0.590 | $0.790 | — | OpenRouter ↗ | API-measured |
| MiniMax M2.7 | $0.600 | $1.80 | — | OpenRouter ↗ | API-measured |
| qwen3.8-27b | $0.800 | $4.00 | — | Pricebook (LiteLLM) ↗ | Official list |