All promotions
GuidesMCP

Groq

Inference providerUS Westprices checked Oct 23, 06:00 UTC

Pricing page ↗

Groq serves 7 models at a verified price. On the 6 it shares with other hosts, it is never the cheapest and charges a median 220% above it.

Models served77 with open weights
Cheapest input, per M$0.05Llama 3.1 8B
Uptime, last 24 h99.5%average of its 6 OpenRouter endpoints
Sign-upno free credits stated
Newest first
ModelPrecisionContextToolsInputOutputCache
Qwen3.8 27Bnot stated by the host128K—$0.80$4.00$0.40
MiniMax M2.7not stated by the host192K✓$0.60$1.80—
gpt-oss-safeguard-20bnot stated by the host128K✓$0.075$0.30$0.0375
gpt-oss-120bnot stated by the host128K✓$0.15$0.60$0.075
gpt-oss-20bnot stated by the host128K✓$0.075$0.30$0.0375
Llama 3.3 70Bnot stated by the host128K✓$0.59$0.79$0.29
Llama 3.1 8Bnot stated by the host128K✓$0.05$0.08$0.025

7 of 7 models · prices per million tokens · precision and context: read at Groq's own catalogue or docs where its terms allow, otherwise as declared to OpenRouter (hover a cell for its source) · tools as Groq declares them to OpenRouter

ConditionWhat it saysSource
Sign-upNo credit card needed to startconsole.groq.com
Free creditsNone stated—
Where data goesRetained customer data stored in the United Statesconsole.groq.com
Headquarters regionUS West—
Price listWhere its per-token prices are publishedgroq.com
API documentationIts own docsconsole.groq.com
Status pageOutages and incidents, as Groq reports themgroqstatus.com
Terms of serviceIts termsconsole.groq.com
No cc requiredFree tier, no credit cardconsole.groq.com

Links checked Oct 8.