Datagrom AI News Logo

Cerebras Powers Perplexity Sonar with Industry’s Fastest AI Inference

Cerebras Powers Perplexity Sonar with Industry’s Fastest AI Inference

February 13, 2025: Cerebras Boosts Perplexity Sonars AI Speed - Cerebras Systems has teamed up with Perplexity to power Sonar, an advanced AI search model based on Llama 3.3 70B, providing unmatched answer quality and speed. Leveraging Cerebras' state-of-the-art AI inference infrastructure, Sonar processes 1,200 tokens per second, ensuring near-instant, precise search results. This partnership signifies a revolutionary shift in search technology, enhancing user experience with real-time, accurate information access.

The Cerebras CS-3 system, featuring the Wafer-Scale Engine-3, supports this high-speed performance, driving innovation in AI applications. Through its pioneering technology, available via Cerebras Cloud and on-premise solutions, the collaboration with Perplexity pushes the boundaries of what AI search models can achieve.

Link to article Share on LinkedIn

Stay Current on AI in Minutes Weekly

Cut through the AI noise - Get only the top stories and insights curated by experts.

One concise email per week. Unsubscribe anytime.