As the year of agents gains momentum, Sana AI's agentic capabilities continue to advance.

February's hero release was Pro, our most advanced mode for AI agents. We coupled Pro with Filters to give you even more control over your agents' data sources. On Monday, Anthropic released Claude 3.7 Sonnet—the first hybrid reasoning model on the market. We made it available within 24 hours.

Pro goes beyond simple queries with multi-step reasoning. It plans, runs dozens of searches across all your internal apps, and processes vast amounts of information to help you deliver hours of work in minutes. You can see its thought process on the right sidebar as you chat.
​
This advanced agent mode uses 30% more sources for higher accuracy, allowing for better synthesis of internal data and public knowledge. Whether you’re working with hundreds of documents or analyzing giant datasets, Pro mode helps you get things done quickly. It is available for Team and Enterprise users.

​
Filters let you guide agents to use specific sources. You can filter on individual assets, time ranges, data types, collections, and integrations for faster and more accurate answers. You can also toggle web access to restrict where your agent looks for information.

Anthropic's most intelligent model to date, already available on Sana AI. Claude 3.7 Sonnet can provide near-instant responses or extended, step-by-step thinking. It's a powerful choice for building work agents focused on solving real-world tasks.