Groq pricing & review
Custom LPU silicon serving open models at hundreds of tokens/second: the speed king for latency-sensitive apps, with simple per-token pricing.
Where Groq wins
- Fastest tokens/sec in the market
- Simple pricing
- Generous free tier for prototyping
Groq model pricing
List prices refreshed 2026-10-08 ยท cheapest 25 shown| Model | Input $/1M | Output $/1M | Market floor in |
|---|---|---|---|
| GPT-OSS 20B | $0.075 | $0.300 | $0.015at Darkbloom |
| GPT-OSS-Safeguard-20b | $0.075 | $0.300 | $0.070at AWS Bedrock |
| GPT-OSS 120B | $0.150 | $0.600 | $0.030at W&B Inference |
| Llama-Guard-3-8B | $0.200 | $0.200 | $0.020at Nebius |
| Qwen3.8 27B | $0.800 | $4.00 | $0.400at W&B Inference |