Groq
79.0Trending upGroq|Rank #53
Ultra-fast AI inference platform powered by proprietary LPU chips delivering the lowest latency for open-source models.
80.0
Performance
42.0
Popularity
82.0
Value
59.3
Trust
Overview
Groq is an AI infrastructure company founded in 2016 by former Google TPU engineers. Its proprietary Language Processing Unit (LPU) delivers sub-300ms time-to-first-token latency, dramatically faster than GPU-based alternatives. GroqCloud hosts open-source models like Llama, DeepSeek, Qwen, and Mixtral via an OpenAI-compatible API. With over 2 million developers, Nvidia agreed to acquire Groq for approximately $20 billion in December 2025.
Pricing Plans
Free
Free
- Basic rate limits
- All hosted models
Developer
Custom
- Pay-per-token
- 10x rate limits
- Batch API (25% off)
Enterprise
Custom
- Custom rate limits
- Fine-tuned models
- SLA
Features
Sub-300ms latency (LPU)
OpenAI-compatible API
Dozens of open-source models
Batch API
Compound models with tools
Quick Info
- Category
- Open Source
- Status
- active
- Monthly Visits
- 1.2M
- API Available
- Yes
- Open Source
- No
