Groq

Groq runs popular open models on its LPU inference stack with very low latency and high throughput via a developer API—ideal for realtime chat, agents, and fast product prototypes.

• Ultra-low latency
• LPU inference
• Developer API