Groq AI models — pricing & benchmarks

Official siteDocsPricingAPI

Groq is the premier neocloud for fast inference. One fully integrated platform for infrastructure, inference, and control. Millions of developers run trillions of tokens on Groq every week. Inference platform · OpenAI-compatible API · Fast Inference · Low Latency · Lpu · Open Weight

Intelligence vs Price

Best value among Groq models on this chart: Qwen3.8 27B · Qwen3.6 27B · GPT OSS 120B (and 2 more on the dashed frontier). Prices use each model's lowest available Groq price across regions. Hover any dot for full pricing, or click a creator in the legend to isolate.

Groq models

20 models in Global, 15 with pricing

All Model Types

All Creators

US Dollar ($)

Per 1M tokens

Model	Creator	Input Price, $	Output Price, $	Context	Max Output	Intelligence	Coding
Qwen3.8 27B	Alibaba	0.8	4.00	1.0M	131K	#1	#1
Qwen3.6 27B	Alibaba	0.6	3.00	262K	236K	#2	#2
GPT OSS 120B	OpenAI	0.15	0.6	131K	131K	#3	#3
GPT OSS 20B	OpenAI	0.075	0.3	131K	131K	#4	#4
Qwen3 32B	Alibaba	0.29	0.59	131K	41K	#5	#5
Llama 3.3 70B Instruct	Meta	0.59	0.79	131K	120K	#6	#6
Llama 3.1 8B Instruct	Meta	0.05	0.08	200K	128K	#7	#7
Gemma 7B IT	Google	0.05	0.08	8K	8K	N/A	N/A
GPT OSS 20B Safeguard	OpenAI	0.075	0.3	131K	66K	N/A	N/A
Kimi K2 Instruct	Moonshot AI (Kimi)	1.00	3.00	262K	100K	N/A	N/A
Llama 4 17B Maverick Instruct	Meta	0.2	0.6	1.0M	16K	N/A	N/A
Llama 4 17B Scout Instruct	Meta	0.11	0.34	10.0M	16K	N/A	N/A
Llama Prompt Guard 2 22M	Meta	0.03	0.03	512	512	N/A	N/A
Llama Prompt Guard 2 86M	Meta	0.04	0.04	512	512	N/A	N/A
LlamaGuard 4 12B	Meta	0.2	0.2	1.0M	16K	N/A	N/A
Orpheus 1 English	Canopylabs	N/A	N/A	4K	N/A	N/A	N/A
Orpheus Arabic Saudi	Canopylabs	N/A	N/A	4K	N/A	N/A	N/A
PlayAI TTS	PlayAI	N/A	N/A	10K	10K	N/A	N/A
Whisper 3 Large	OpenAI	N/A	N/A	N/A	N/A	N/A	N/A
Whisper 3 Large Turbo	OpenAI	N/A	N/A	N/A	N/A	N/A	N/A

Groq

Intelligence vs Price

Groq modelsAPIGET/api/v1/providers/groq/models

Groq models