Categories / Groq

Fast AI inference cloud (GroqCloud) with an OpenAI-compatible API for open models and Whisper speech-to-text. The free tier is for testing; the Developer tier is pay-as-you-go per token.

Free plan
Yes
Starting price
Paid price not found
Pricing model
Freemium
Verification
Verified
Last verified

Prices may change. These are correct as of 10 Nov 2026. Please confirm on the vendor's own page before you pay.

Verified. Groq's docs publish per-token rates and no monthly plan price, so we record the rates in the plan notes. Some larger models are listed as 'contact sales'.

Free plan

Free plan. Free tier for building and testing with lower rate limits and community support.

Plans and prices (USD)

Swipe the table sideways to see what each plan includes.

PlanBilled monthlyYearly, per monthWhat it includes
Free
per organization
$0 not recorded
  • Build and test on Groq, community support
  • 25 MB audio file limit
Developer
per usage
not recorded not recorded
  • Pay per token, billed at month end or at $1, $10, $100, $500, $1,000 thresholds
  • GPT OSS 120B $0.15 input / $0.60 output per 1M tokens; GPT OSS 20B $0.075 / $0.30
  • Whisper large v3 $0.111 per hour; Whisper large v3 turbo $0.04 per hour
  • Batch API is 50% cheaper; 100 MB audio limit
Enterprise
per custom
not recorded not recorded
  • Scalable capacity, SSO, 90-day audit logs, LoRA inference

"Not recorded" means we did not find that price, not that it is free. Prices exclude tax and can differ by country.

Where it may fit

  • D
  • e
  • v
  • e
  • l
  • o
  • p
  • e
  • r
  • s
  • w
  • h
  • o
  • n
  • e
  • e
  • d
  • v
  • e
  • r
  • y
  • f
  • a
  • s
  • t
  • o
  • p
  • e
  • n
  • -
  • m
  • o
  • d
  • e
  • l
  • i
  • n
  • f
  • e
  • r
  • e
  • n
  • c
  • e
  • b
  • y
  • A
  • P
  • I
  • .

Taken from what the vendor says the tool does. We have not tested it.

What to watch

  • Rates change as models are added or retired.
  • Some Llama models are listed as 'contact sales'.

Sources

Open Groq's own site · Report a mistake on this page