Groq

Groq is an AI inference company whose GroqCloud service runs open-weight models on its own hardware rather than routing requests to other providers. That hardware is the LPU (Language Processing Unit), a chip designed specifically for running language models rather than training them, which keeps time-to-first-token low and generation speed high and consistent from one request to the next.

Groq's model lineup is curated to what runs well on LPUs: Meta's Llama family, OpenAI's open-weight gpt-oss models, Moonshot's Kimi, Alibaba's Qwen, and speech models such as Whisper, all behind an OpenAI-compatible API, so existing SDKs work by changing the base URL. Pricing is per token with no subscription and among the lowest of the major providers for small and mid-size models. A free tier with rate limits covers experimentation, and the Batch API and prompt caching each halve the cost of eligible workloads.

In December 2025, Nvidia signed a non-exclusive license to Groq's inference technology, reported at about $20 billion, and hired its founder and several senior leaders. Groq continues as an independent company and GroqCloud keeps operating, so the buying question is unchanged: for real-time chat, voice agents, and coding assistants where latency shapes the experience, Groq is usually among the fastest options at its price point.