Groq API pricing

Checked 12 Sept 2026 LLMs

12 models from US. Blended cost runs from $0.058 to $1.60 per million tokens. LLM vendors run no affiliate programmes, so nothing on this page earns Pikkly anything.

Every model

Ranked by blended cost at 3:1 input to output.

Calculator →
Groq models by blended cost per million tokens — checked 12 Sept 2026
Source
gemma-7b-it Best value $0.050 $0.080 8,192 $0.058
gpt-oss-20b $0.075 $0.300 $0.037 131,072 $0.131
gpt-oss-safeguard-20b $0.075 $0.300 $0.037 131,072 $0.131
gpt-oss-120b $0.150 $0.600 $0.075 131,072 $0.263
qwen3-32b $0.290 $0.590 131,000 $0.365
qwen3.6-27b $0.600 $3.00 131,072 $1.20
kimi-k2-instruct-0905 $1.00 $3.00 $0.500 262,144 $1.50
qwen3.8-27b $0.800 $4.00 131,042 $1.60

Last checked . Methodology

All models · llama-3.1-8b-instant · llama-4-scout-17b-16e-instruct · llama-4-maverick-17b-128e-instruct · llama-3.3-70b-versatile · Run your own on a GPU