Google's October Gemini update adds 11 features, from a Windows app to shareable avatars

What Google released

On Friday, October 30, 2026, Google published the October edition of Gemini Drops, its regular roundup of new features for the Gemini app. As jetstream.blog reported, this month's list has 11 items spread across the phone and web app, desktop computers and Google Maps.

The list covers a new AI model called Gemini 4 Argon, a Mac app with a coding tool attached, a Windows app, a saved-prompt feature called Skills, a Maps upgrade for Personal Intelligence, a link to the Gemini Notebook research tool, more App Connectors, file creation and editing, an interactive learning tool, the Lyria 3.5 music model, and shareable avatars.

Not everything is new this month. Some items, like the Windows app and Lyria 3.5, launched in September and are included in the roundup.

A new model and two desktop apps

Gemini 4 Argon is the first model in the Gemini 4 family, which Google announced in late September. (A model is the AI system that does the thinking behind the chat.) It became generally available worldwide on October 13. Google describes it as strong at programming, finance, legal work and cybersecurity defense, and says it breaks complex tasks into steps and can produce things like web apps, 3D files and slide decks.

You can use it in the Gemini app and in AI mode in Search if you subscribe to Google AI Pro or Google AI Ultra. It appears as an extra choice in the model list and does not replace 3.1 Pro.

The Windows Gemini app arrived on September 10 for PCs running Windows 10 or later. Pressing Alt + Space summons it over whatever you are doing, for example to fact-check a document or brainstorm slide titles. It also reaches Gemini Spark (Google's personal AI agent, meaning an assistant that carries out tasks), Nano Banana for images and Gemini Omni for video.

On the Mac, the Gemini app now has an Antigravity button. Antigravity is Google's AI coding environment. After you approve its plan, it builds and tests software in the background and tells you when there is something to review. For now it works in English for users 18 and older, in regions where Antigravity is supported.

Features that draw on your own stuff

Skills began rolling out on October 13. A skill is a saved set of instructions. Type a slash and its name to run it, or let Gemini apply enabled skills on its own when your request fits. You can combine several at once and manage them on a Skills page. Gems, Gemini's custom AI helpers, are being phased out, and automatic migration to Skills starts on November 17, 2026.

Personal Intelligence, Gemini's personalization feature, now includes Google Maps among the apps it can read. It can use your place search history and saved locations. Google's example asks Gemini to name a matcha soft-serve shop from last year's Kyoto trip and say whether it is open now, pulling from Google Photos and Maps. The Maps piece is rolling out gradually.

Gemini Notebook, the AI research tool, can now be reached from a Gemini chat by typing @Gemini Notebook. You can add material or ask questions about what is inside. The link is on by default for personal and Workspace accounts, but off by default in the European Economic Area.

Connectors, the earlier App Connectors renamed in mid-October, gained Adobe, Strava and Peloton. Gemini can now link to over 40 apps and services. Most are available in US English first.

Making files, learning and playing

Since October 29, Gemini can create documents and PDFs inside the chat, then let you preview and edit them there. You can type changes in full-screen view or highlight text and ask Gemini to rework it. Google's example turns a photo of handwritten lecture notes into a PDF study guide. PDF editing works on web, Android and iOS, while Docs, Sheets, Slides, Markdown, text and code files are web-only for now.

A student-focused tool called Step inside with immersive explorations began rolling out on October 28 in the students hub. It lets you tap labels to move from a large scale, like a planet, down to molecules and atoms, with topics such as the Silk Road or a jet engine. Audio narration and a shortcut from chat are promised soon.

Lyria 3.5, Google's latest music model, lets you pick vocals or instrumental, short or long tracks, and templates such as birthday songs. Avatar sharing lets you give friends or family access to your avatar, then tag them in prompts to make images and videos of you together. You can remove access at any time.

Where it works and where it doesn't

Availability is the catch. Of the 11 features, seven are available in Japan: Gemini 4 Argon, the Windows app, Skills, Maps in Personal Intelligence, Notebook integration, file editing and Lyria 3.5. The other four are not: the macOS Antigravity link, expanded Connectors, immersive explorations and avatar sharing.

Other limits apply too. Immersive explorations supports English, Indonesian, Hindi, Brazilian Portuguese, French and Simplified Chinese, so a user in Japan can use it in English. Avatar sharing is rolling out in English outside the EEA, Switzerland and the UK. Gemini 4 Argon needs a paid plan. The roundup does not state pricing for the other features.