Google launches Deep Research and Deep Research Max agents to automate complex research

In Google's own benchmarks, Deep Research Max shows a big jump on retrieval and reasoning tasks. The agent pulls from more sources than the previous version and catches nuances the older model tended to miss, Google says.

The comparison with OpenAI's GPT-5.4 and Anthropic's Opus 4.6 isn't quite apples to apples. GPT-5.4 is great at autonomous web search, but it isn't tuned for deep research. For that, OpenAI ships its own DR agent, which switched to GPT-5.2 after the February update, not GPT-5.4. OpenAI's strongest search model is actually GPT-5.4 Pro, which Google apparently left out of the comparison. According to OpenAI, GPT-5.4 Pro hits up to 89.3 percent on the agentic search benchmark BrowseComp, while GPT-5.4 lands at 82.7 percent.

Anthropic also reports higher BrowseComp numbers for Opus 4.6 than Google shows, specifically 84 percent. Anthropic says the model hit that score with reasoning turned off, since it actually performed better that way than at the high reasoning intensity Google used in its API benchmark.

The gaps likely come down to testing methodology, whether the models were evaluated through the raw API or wrapped in each lab's own tooling. Google's numbers aren't necessarily wrong, but they're worth reading with some caution. Either way, the presentation isn't transparent.

MCP support opens the agent to proprietary data

One of the bigger changes is support for the Model Context Protocol (MCP). Developers can wire Deep Research into their own data sources and specialized feeds, like financial or market data providers. By accepting any tool definition, the agent shifts from a pure web searcher to a full autonomous agent that can query specialized databases, Google says.

For the first time in the Gemini API, the agent can also generate native charts and infographics directly inside reports, rendered either as HTML or in "Nano Banana" format, making it easier to present complex data visually.

Other additions include collaborative planning, which lets users review and tweak the agent's search plan before it runs, multimodal input from PDFs, CSVs, images, audio, and video, and real-time streaming of intermediate steps. Developers can also shut off web access entirely and limit the agent to their own data.

Google says the agents run on the same infrastructure that powers research features in its consumer products, including the Gemini app, NotebookLM, Google Search, and Google Finance. Developers can start building custom search workflows through the Interactions API, and both agents will roll out to startups and enterprise customers through Google Cloud soon.