Google DeepMind has rolled out Deep Research Max, an AI agent built on Gemini 3.1 Pro that autonomously conducts research across web and proprietary data sources. Developers can now integrate custom data feeds, including financial information, through the Model Context Protocol.
Google DeepMind's Deep Research Max represents a significant expansion of the company's AI research capabilities. The agent operates on Gemini 3.1 Pro and automates the process of conducting complex research by autonomously navigating both public web sources and proprietary databases.
A key feature of Deep Research Max is its integration with the Model Context Protocol, which allows developers to connect specialized data sources directly to the agent. Financial feeds and other industry-specific information can now be plugged into the system, enabling more targeted and comprehensive research workflows.
The launch addresses growing demand for automation in research-heavy workflows. Rather than manual web searches and data compilation, organizations can deploy these agents to gather, synthesize, and analyze information across multiple sources simultaneously.
Deep Research and Deep Research Max sit within Google's broader push into AI agents. The distinction between the two versions suggests a tiered offering, with Deep Research Max likely providing enhanced capabilities or longer research runs.
Google has released benchmark data alongside the announcement, though the company maintains limited transparency around the evaluation metrics used. This approach mirrors Google's previous AI releases, where benchmark details often remain proprietary.
The timing aligns with increased competition in AI research automation. Other AI labs have similarly pursued agent-based solutions for reducing manual research work, making this space increasingly crowded.
Developers can integrate Deep Research Max into their applications through APIs, with the Model Context Protocol serving as the primary integration point for custom data sources. The exact pricing and availability details for enterprise deployments remain to be confirmed.
Google DeepMind has not announced significant limitations on research scope or frequency, though such constraints typically apply to autonomous agent systems managing external data access.
PrismML has compressed a 27-billion-parameter AI model to under 4 GB, enabling it to run directly on iPhone devices. The compressed model retains 90 percent of its original performance with minimal impact on math and coding capabilities.
Israel-based Hemispheric secured $52 million in funding for its AI model that analyzes non-invasive brain activity measurements and converts them into quantitative diagnostic metrics.
Anthropic and Blackstone are backing Ode, a new venture that embeds AI engineers directly inside enterprises. The bet signals a shift in where the next trillion dollars in AI value may be created: not in building models, but in implementing them.
Spectro Cloud, an AI infrastructure company focused on managing token costs, secured $100 million in Series D funding at a valuation exceeding $1 billion. The raise marks significant growth from the company's $750 million valuation in 2024.