Google has launched two new autonomous research agents, Deep Research and Deep Research Max, available through the paid tiers of the Gemini API. The new agents, built on the Gemini 3.1 Pro model, replace a preview version released in December and are designed to automate complex research and analysis.
The standard Deep Research agent is optimized for speed and lower cost in real-time applications, while the Max version uses more computation for deeper, more thorough analysis in asynchronous tasks. For the first time, these agents can generate native charts and infographics and can connect to proprietary data sources, like financial data streams, in addition to the open web. This move positions Google to compete more directly with enterprise AI offerings from rivals like Microsoft.