Perplexity Pro gives you access to the latest AI models from OpenAI and Anthropic, as well as other companies and includes both proprietary and open sourced models.

Here’s a quick overview of each model currently included—the list evolves as models and search modes update.

"Best" is optimized for speed, web browsing and taking the decision-making out of your hands when it comes to selecting a model. It’s fine-tuned to deliver fast, accurate answers and will automatically select the most suitable model for your query.

Pro search uses your default or baseline model and performs an in depth internet search, granting you 3x the amount of sources. You can pick an advanced model when performing a Pro Search of your choice. The latest advanced models we support are:

This model is designed for complex problem-solving across domains, with a strong focus on coding, instruction following, and handling large context windows.

Claude 3.7 Sonnet

3.7 Sonnet is one of the most advanced large language model (LLM) available, particularly excelling at handling nuanced language tasks with a strong balance of speed and accuracy.

Built on the LlaMa 3.1 70B open-source model, Sonar is trained in-house to work seamlessly with Perplexity’s search engine, making it a perfect fit for our cutting-edge search tech.

Grok 3 Beta, the latest advanced reasoning model from xAI, excels in complex math, science and coding tasks.

Google’s latest and most advanced generative AI model, introduced in March 2025, represents a significant leap in AI reasoning, coding, and multimodal capabilities

These models can be picked when using Pro Search. We currently offer two models.

o4-mini introduces significant improvements over its predecessor, o3-mini. It offers more accurate and reliable responses to complex prompts, especially those requiring precise adherence to instructions. Pro subscribers can try o4-mini on all platforms via the model selector.

DeepSeek-R1 redefines AI reasoning through cost-efficient, open-source innovation, specializing in logical inference, mathematical problem-solving, and real-time decision-making.

To learn more about the r1 reasoning model, visit our specialized FAQ.

3.7 Sonnet is one of the most advanced large language model (LLM) available, particularly excelling at handling nuanced language tasks with a strong balance of speed and accuracy.

The best way to figure out which model works for you is to test them out. Try comparing your queries across different models and see which one suits your needs best. It all comes down to what you’re looking to achieve.

A token is a small chunk of text that AI models break down to process. Think of tokens as building blocks that help AI understand and generate language. You can use a tokenizer to see how many tokens a given piece of text contains.

No, you can use these models just as you would in their original platforms.

While all up to date LLMs for the most part achieve similar performance on standard benchmarks, Perplexity's choice of models have distinct operational strengths optimized for web search use case.