Gemini Deep Research gets Max tier

This is Google turning Deep Research from a Gemini feature into infrastructure developers can build serious analyst workflows around.

• –MCP support is the biggest developer hook: teams can connect financial, market, file, and internal data sources instead of relying only on public web search
• –Deep Research Max is aimed at asynchronous enterprise jobs, where spending more test-time compute for better synthesis makes sense
• –Native visual output matters because research agents that produce stakeholder-ready charts and infographics reduce a major handoff step
• –The risk remains factual reliability: long-running agents compound small retrieval or reasoning errors, so citations, source diversity, and human review still matter
• –This puts Google more directly against OpenAI, Perplexity, Genspark, and other research-agent stacks, but with Google-scale search and Gemini API distribution