Gemini in Chrome (Google)
Google’s first-party announcement (blog.google) of Gemini in Chrome — its in-browser AI assistant — expanding to Latin America, Africa, the Middle East and more (desktop + iOS). Tier T3 (first-party vendor announcement; feature/geography claims are a volatile snapshot). Ingested here as the incumbent’s move in the browser AI race and the anchor for the assistant pole of ai-browser.
What it does
- Cross-tab reasoning — summarize and compare across open tabs.
- App integrations — Calendar scheduling, Maps, Gmail drafting, YouTube Q&A.
- In-side-panel image editing (“Nano Banana 2”).
- Conversation memory, and a premium “Personal Intelligence” that ties Gmail/Photos/YouTube/Search into tailored answers.
Why it’s here — the assistant pole
Google positions this explicitly as an assistant, not autonomous browsing: it answers, summarizes, and drafts, but ships prompt-injection protections and confirm-before-sensitive- action gates rather than acting unattended. That places it opposite the agentic pole (Comet / Atlas / Opera Neon, which act to complete tasks) — the central assistant ↔ agent axis of this spoke. It is also the incumbent angle: Gemini-in-Chrome is AI added inside the dominant browser, versus the challengers building AI-first browsers from the tab up.
Cross-spoke
The underlying model is gemini (llm-providers/research own the model market); the “AI answers in the browser” angle is the browser-side of ai-search-shift (search-marketing). This page keeps the browser-product substance.