Llama 3 Sonar Online is an 8x7B parameter model based on Meta's Llama 3.1 model. It is fine-tuned for chat completions and served by Perplexity. The model has access to recent internet knowledge when forming responses.

Gemini 2.0 Flash delivers next-gen features and improved capabilities, including superior speed, native tool use, multimodal generation, and a 1M token context window.