Your cookie choices

We use necessary cookies to make this site work. With your permission, analytics cookies help us understand usage and marketing cookies support advertising.

You can reject non-essential cookies and still use the site. Change your choices anytime in Cookie settings.

Cookie policy
Cookie settings

Groq review

Updated September 28, 2026 · DeployCue Editorial Team · Facts checked September 26, 2026

We earn a commission when you sign up through links on this site. Partner status can affect where a brand appears in our rankings.

Our verdict

Groq, an OpenAI-compatible API run by Groq LLC in the US, suits developers who want cheap inference on a fixed set of open-weight models. Input fees run $0.03 to $0.80 and output fees $0.03 to $4, with a free tier and occasional promotional credits, while the catalogue stays narrow.

Key facts

Best for
Low cost OpenAI-compatible inference on a short, fixed list of open-weight models.
Input price from
$0.030 to $0.80 (Source, checked September 26, 2026)
Output price from
$0.030 to $4 (Source, checked September 26, 2026)
Models offered
Llama-3.1-8b-instant, Llama-3.3-70b-versatile, Openai/GPT-OSS-120b, Openai/GPT-OSS-20b, Openai/GPT-OSS-safeguard-20b, Qwen/qwen3.8-27b, Minimaxai/minimax-m2.7, Meta-llama/llama-prompt-guard-2-22m, Meta-llama/llama-prompt-guard-2-86m, Whisper-large-v3, Whisper-large-v3-turbo, Canopylabs/orpheus-v1-english, Canopylabs/orpheus-arabic-saudi (Source, checked September 26, 2026)
Largest context window
196,608 (Source, checked September 26, 2026)
Free credits
Occasional promotional credits; free tier available (Source, checked September 26, 2026)
OpenAI-compatible API
Yes (Source, checked September 26, 2026)
Legal entity
Groq LLC (Source, checked September 26, 2026)
Headquarters
United States (Source, checked September 26, 2026)
Support channels
Email, Chat, Help-centre, Community (Source, checked September 26, 2026)
Official website
groq.com

Apps and profiles

Elsewhere
WikidataWikipediaLinkedInX

Pricing

  • Groq Qwen3 32B (USD)

    input-tokens: $0.29 / million

  • Groq Qwen3 32B (USD)

    output-tokens: $0.59 / million

  • Groq Mixtral 8x7B (USD)

    input-tokens: $0.24 / million

  • Groq Mixtral 8x7B (USD)

    output-tokens: $0.24 / million

  • Groq Llama 3.3 70B (USD)

    input-tokens: $0.59 / million

  • Groq Llama 3.3 70B (USD)

    output-tokens: $0.79 / million

  • Groq Llama 3.1 8B (USD)

    input-tokens: $0.05 / million

  • Groq Llama 3.1 8B (USD)

    output-tokens: $0.08 / million

  • Groq Kimi K2 (USD)

    input-tokens: $1.00 / million

  • Groq Kimi K2 (USD)

    output-tokens: $3.00 / million

  • Groq Kimi K2 (USD)

    cached-input-tokens: $0.50 / million

  • Groq gpt-oss-20b (USD)

    input-tokens: $0.08 / million

  • Groq gpt-oss-20b (USD)

    output-tokens: $0.30 / million

  • Groq gpt-oss-20b (USD)

    cached-input-tokens: $0.04 / million

  • Groq gpt-oss-120b (USD)

    input-tokens: $0.15 / million

  • Groq gpt-oss-120b (USD)

    output-tokens: $0.60 / million

  • Groq gpt-oss-120b (USD)

    cached-input-tokens: $0.08 / million

  • Groq Gemma 2 9B (USD)

    input-tokens: $0.20 / million

  • Groq Gemma 2 9B (USD)

    output-tokens: $0.20 / million

  • Groq DeepSeek R1 Distill (USD)

    input-tokens: $0.75 / million

  • Groq DeepSeek R1 Distill (USD)

    output-tokens: $0.99 / million

Pros and cons

Pros

  • Input fees run $0.03 to $0.80 and output fees $0.03 to $4, a low starting range for chat models.
  • The API is OpenAI-compatible, so most client code carries over as it is.
  • A free tier and occasional promotional credits cover trials.

Cons

  • The catalogue is fixed at llama 3.1 8b instant, llama 3.3 70b versatile and openai/gpt oss 120b, so your own model cannot be added.
  • The largest context window is 196,608, which places Groq 59 of 72 here.
  • Support runs on email, chat and help centre only, with no dedicated phone line listed.
Groq

Ready to try Groq?

Scored 9.4 / 10 by DeployCue.

See alternatives

Who it suits

  • Developers who want an OpenAI-compatible endpoint and already know which open-weight model they need.
  • Teams prototyping on the free tier before paying for usage.
  • Product teams adding speech to text or prompt guard checks through the same endpoint.

Who should look elsewhere

  • Teams needing a long context window, since Groq stops at 196,608 while the site median is 1,000,000.
  • Buyers who want a wide hosted catalogue: Azure AI Foundry lists many more model families and sizes.
  • Shoppers chasing the lowest entry input price alone, because Salad quotes input from $0.06.

Plans, egress and billing

Costs differ by model rather than by a single plan. Input fees run $0.03 to $0.80 and output fees $0.03 to $4, so a small model sits at one end of that range and a larger one at the other. A free tier is available, plus promotional credits from time to time, which covers trials and light use.

Core features

The API is OpenAI-compatible, so existing client code usually carries over unchanged. The catalogue in llama 3.1 8b instant, llama 3.3 70b versatile and openai/gpt oss 120b mixes chat models, Whisper speech to text and prompt guard models. The largest context window is 196,608, enough for long chat histories but short beside the biggest windows in this group.

How Groq compares

Frequently asked questions

How much does it cost to run a model on Groq?+

Input fees run $0.03 to $0.80 and output fees $0.03 to $4, and the model you choose sets where you land in those ranges. Groq sits 18 of 71 on entry input price and 6 of 70 on entry output price.

Is there a free tier to try?+

A free tier is available and promotional credits turn up from time to time, so small tests can run before any spend. Once usage is metered, the bill follows the model: input fees of $0.03 to $0.80 and output fees of $0.03 to $4. Prototypes sit at the entry end of that range.

Do I need to rewrite my code to use Groq?+

The API is OpenAI-compatible, so OpenAI client code can point at Groq through a base URL change. The model names you request come from Groq's own list in llama 3.1 8b instant, llama 3.3 70b versatile and openai/gpt oss 120b, so the identifiers differ.

How large a context window does Groq offer?+

The largest context window across the catalogue is 196,608, against a site median of 1,000,000. That puts Groq 59 of 72 providers here, so very long documents are better chunked or handled by a gateway with longer limits.

Which models can I call?+

The list in llama 3.1 8b instant, llama 3.3 70b versatile and openai/gpt oss 120b covers Llama and Qwen chat models, Whisper speech to text, prompt guard models and English and Arabic text to speech voices. It is a fixed set, so check your model appears there.

Who is behind Groq and how do I get help?+

The API is operated by Groq LLC, a company headquartered in the US. Support and billing questions run through email, chat and help centre, with no dedicated phone line for account problems. Email and the help centre are the broadest of those routes.

Sources

  • console.groq.com/docs/billing-faqs checked September 26, 2026

    Backs: Free credits, Support channels

    “Yes! We occasionally offer promotional credits, such as during hackathons and special events.”

  • console.groq.com/docs/models checked September 26, 2026

    Backs: Largest context window, Input price from, Models offered, Output price from

    “196,608”

  • console.groq.com/docs/openai checked September 26, 2026

    Backs: OpenAI-compatible API

    “We designed Groq API to be mostly compatible with OpenAI's client libraries, making it easy to configure your existing applications to run on Groq”

  • groq.com/terms-of-use checked September 26, 2026

    Backs: Headquarters, Legal entity

    “Groq LLC, Attn: Legal Department, P.O. Box 1778, Mountain View, CA, 94042.”

Compiled from 4 source pages, checked September 26, 2026. We did not open an account.

Groq

Groq

Low cost OpenAI-compatible inference on a short, fixed list of open-weight models.

9.4/10See alternatives