Google DeepMind Launches Deep Research Max: Agent That Runs 160 Searches While You Sleep

Google DeepMind has just released in public preview two autonomous research agents that promise to change how scientists and professionals conduct investigation: Deep Research and Deep Research Max.

How It Works

Built on Gemini 3.1, these agents can:

• Execute up to 160 parallel searches while the user sleeps
• Synthesize information from multiple sources
• Deliver a complete, structured report by morning
• Autonomously refine research questions

The Autonomy Leap

Unlike traditional search tools, Deep Research Max doesn't just collect information - it understands the research context, identifies knowledge gaps, seeks additional sources to fill them, and organizes results into a coherent narrative.

This represents a significant leap in AI agent autonomy for research and analysis. Instead of spending hours manually collecting information, the researcher defines the initial scope and the agent does the heavy lifting.

Availability

Both agents are available in public preview via the Gemini API. Deep Research Max is the more robust version, with extended execution capacity and greater investigation depth.

For professionals who depend on intensive research - journalists, academics, market analysts, and developers - this tool could be a productivity game-changer.