Sonar Pro

Sonar Pro is Perplexity's flagship model in the Sonar family, designed to combine traditional language model capabilities with real-time web search and information retrieval. Unlike standard language models that rely solely on training data, Sonar Pro can access current web information to provide up-to-date responses and citations. The model supports both text and image inputs with a 200,000 token context window, allowing it to process lengthy documents while incorporating visual content analysis. Sonar Pro integrates search capabilities directly into its inference process, enabling it to retrieve, synthesize, and cite current information from the web rather than being limited to a static knowledge cutoff. This search-augmented approach makes Sonar Pro particularly valuable for applications requiring current information, fact-checking, or research assistance. It competes with other search-enhanced AI systems by providing cited responses that users can verify, though it operates differently from traditional language models that generate responses purely from learned parameters.