# Groq

> A company with its own fast AI chips. It runs open models for you.

**Field:** Artificial Intelligence  
**Category:** Models & Providers  
**Also known as:** GroqCloud, Groq LPU, LPU, Language Processing Unit

## Think of it like

It is like a race car built for one road. It does one job, very fast.

## Definition

Groq is a company that builds chips for running AI models, and a cloud that sells their speed. Its chip is called the LPU, and it is made for answering, not for training. On its cloud, GroqCloud, you pick one of the open models it hosts and send it a prompt. The answers come back very fast. Groq now also pairs its chips with NVIDIA's graphics cards.

## Example

Lucy's voice app must answer before a caller hears a pause. She runs GPT-OSS on Groq, because speed matters most to her.

## Why it matters

Speed changes what you can build. A voice helper or a live game needs an answer in a blink.

## Common mistake

Thinking Groq trains its own chat models. It runs other labs' models on its chips.

## Three principles

- **Own chips.** The LPU.
- **Built for answers.** Inference, not training.
- **Open models.** Hosted on its cloud.

## Related terms

- [Inference Provider](https://learn.benibix.com/term/ai/inference-provider)
- [AI Accelerator](https://learn.benibix.com/term/infrastructure/ai-accelerator)
- [Tokens per Second](https://learn.benibix.com/term/ai/tokens-per-second)
- [Inference (AI)](https://learn.benibix.com/term/ai/inference)
- [gpt-oss](https://learn.benibix.com/term/ai/gpt-oss)

## Told apart from

- [Grok vs Groq](https://learn.benibix.com/compare/grok-vs-groq): People think Grok and Groq are one thing, because the names sound the same. Grok is a chat app and its models, made by SpaceXSI. Groq is a firm that builds fast chips and runs other labs' open models on them. You talk to Grok. You send your own app's prompts to Groq.

## Taught in

- [AI Inference Providers](https://learn.benibix.com/learn/ai-models-providers/ai-inference-providers): AI Models and Providers, 12/13

![Two lanes split by a dashed line. Top lane, model hub: the hub holds model files and cards, and an arrow labelled download carries one to your computer, where you run it; Hugging Face is the example. Bottom lane, inference provider: your app sends a prompt to the provider's chips, an answer comes back, and you pay per token; Groq, Together AI and Replicate are the examples.](https://learn.benibix.com/diagrams/ai2-hub-or-provider.svg)

---

Source: https://learn.benibix.com/term/ai/groq
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-19
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
