Last updated · 114 releases tracked since October 2019
Google Gemini updates and release notes covering Gemini models, the Gemini app, Workspace features and developer tools like Firebase Studio. Entries come from the Google Blog and Firebase Studio release notes, each dated and linked to its source.
Gemini 3.5 Live Voice Features for Google Workspace added voice-activated commands to Gmail, Docs, and Keep. Google Arts & Culture Lab launched AI-powered museum experiences. Gemini 4 Argon introduced a frontier-level AI model for software engineering, enterprise tasks, and cybersecurity.
Summary of September 2026, generated from the releases listed below.
This month
2
-1 vs last month
Last 12 months
94
Busiest: March 2026
Most shipped
Product
55 releases
Newest first. Each entry links to the official page it came from.
Launch of Guided Vision in Gemini Live for Android, offering conversational real-time visual assistance for accessibility, developed with collaboration from the blind and low-vision community.
Asset Studio's multimodal video creation tool powered by Gemini Omni, which turns website URLs and photos into high-quality videos for YouTube ads.
Gemini 4 Argon is a frontier-level AI model designed for complex workloads including software engineering, enterprise knowledge-based tasks, and cybersecurity defense. It features advanced reasoning capabilities, enterprise workflow support, and enhanced security features.
Google Arts & Culture Lab launches two AI-powered interactive experiences: 'The Talking Museum' (Gemini-assisted 3D artifact exploration) and 'Museum Memory' (art memory game).
New voice-activated features for Gmail, Docs, and Keep using Gemini 3.5 Live integration. Allows conversational commands for inbox search, document drafting, and note organization.
Introduction of Expert Intelligence feature allowing integration of trusted sources into Gemini Notebook, enabling users to incorporate insights from leading authors, publications, and domain experts directly into their notebooks.
New 1-year free Google AI Plus plan for students with enhanced features and learning tools
Gemini Spark now integrates with Chrome for enhanced web browsing capabilities, allowing users to handle complex web tasks using logged-in accounts and saved passwords.
Introduces Gemini Robotics 2, a next-generation AI model enabling full-body control, precise manipulation, and multi-robot collaboration for complex tasks.
Gemini Spark, an AI Agent that works 24/7, is now officially available in Thailand for Google AI Pro members.
Nuestro modelo principal que ofrece un mejor rendimiento en codificación, trabajo de conocimiento y multimodalidad. Según el Índice de Análisis Artificial , reduce el uso de tokens de salida en un 17 % en comparación con 3.5 Flash, y en algunos benchmarks como DeepSWE de Datacur…
Workspace creation and user signup disabled as part of Firebase Studio sunset plan. Existing users can access workspaces until March 22, 2027.
New QAT checkpoints for Gemma 4 optimize memory efficiency and on-device performance through advanced quantization techniques.
Meet the new Google Home Speaker, built for Gemini
Gemini Omni is a new model that can create anything from any input, starting with video. It allows combining images, audio, video, and text as input to generate high-quality videos grounded in Gemini's real-world knowledge, with editing capabilities through conversation.
Introducing Universal Commerce Protocol (UCP) features and AI tools to enhance shopping experiences across Google services, including Universal Cart for seamless checkout, AI performance insights for merchants, and expanded UCP integrations for retailers.
A new experiment enhances group meetings on Google Beam with true-to-life size rendering and spatial audio for improved inclusivity and connection.
Gemini API File Search now supports multimodal data, custom metadata, and page-level citations for efficient RAG systems.
Launch of 'Memoria' feature for personalized conversations and tools to import chat history from other AI apps
Launched pronunciation practice tool for Google Translate app on Android, using Gemini AI models for real-time conversation support.
Gemini Enterprise Agent Platform is a developer platform for building, scaling, governing, and optimizing agents, integrated with data and security capabilities.
AI-powered Chrome features including Gemini in Chrome, side panel assistance, Nano Banana 2 image transformation, and enhanced security capabilities
Gemini app introduces Personal Intelligence to create personalized images using Nano Banana and Google Photos, allowing users to generate custom images without long prompts.
Google introduces AI Mode enhancements in Chrome for seamless web exploration, including side-by-side browsing, cross-tab search capabilities, and improved context-aware interactions.
Introduced prepay billing model for Gemini API to provide developers with greater control and predictability over API spend. Allows purchasing API credits upfront with automatic reload options and transparent budget management within Google AI Studio.
The Gemini app is now available as a native macOS experience, offering faster AI assistance directly on desktops with a keyboard shortcut (Option + Space) and free download for macOS 15+ users.
Launch of Personal Intelligence feature in the Gemini app, enabling secure integration with apps like Gmail and Google Photos to provide personalized assistance.
Notebooks in Gemini help organize chats and projects with AI-powered research partner NotebookLM, enabling personal knowledge bases synced across Gemini app and NotebookLM for enhanced workflows.
Released Gemini API Docs MCP and Gemini API Developer Skills to improve coding agents' performance by providing up-to-date API documentation and best-practice instructions.
Google's most cost-effective video generation model, Veo 3.1 Lite, is now available to developers via the Gemini API, offering 50% lower cost than Veo 3.1 Fast with same speed.
Gemini app's Gem feature introduces 'Care Record Assistant' to streamline documentation for caregivers, reducing time spent on record-keeping by 20% through AI-generated SOAP/F-DAR format drafts.
Global expansion of Search Live to all languages and countries where AI mode is available, utilizing the new Gemini 3.1 Flash Live model for multilingual interactions and real-time assistance via voice, camera, and text.
Google launched Gemini 3.1 Flash Live as its highest-quality audio model for lower-latency, more natural real-time voice interactions.
Google added Lyria 3 Pro and wider product support, extending music generation to longer tracks with more control over structure.
Google's Lyria 3 music generation models are now available to developers in public preview, offering two variants for full songs and shorter clips with advanced control features.
Announced sunset of Firebase Studio on March 22, 2027. Recommended migration to Google AI Studio or Google Antigravity.
Firebase Studio announced it will sunset on March 22, 2027 and directed users to migrate projects to Google AI Studio or Google Antigravity. Existing Firebase deployments continue to run after the sunset date.
Introducing the new full-stack vibe coding experience in Google AI Studio with the Antigravity coding agent, Firebase backend integrations, and tools for building real-time apps, databases, and modern web applications.
Expanding Personal Intelligence across AI Mode in Search, the Gemini app, and Gemini in Chrome for free-tier users in the U.S.
Introduces Ask Maps, a conversational experience for real-world questions with personalized recommendations, and Immersive Navigation with 3D visuals, intuitive guidance, and real-time updates for driving.
Expansion of Gemini in Chrome to India, New Zealand, and Canada with new AI features and integrations
Google expanded Gemini in Docs, Sheets, Slides, and Drive with beta features that pull from files and email to draft, edit, and search faster.
New AI features including Circle to Search, Gemini task handling, Magic Cue for restaurant recommendations, and updates to Now Playing and Pixel Watch security.
Google released Gemini 3.1 Flash-Lite in preview for the Gemini API and Vertex AI, targeting high-volume workloads with lower cost and faster responses.
Firebase Studio raised the default output token limit, added Gemini 3.0 Flash Preview and 3.0 Pro Preview, and updated the App Prototyping agent with newer security-rules and workspace behavior.
Google shipped Nano Banana 2, also called Gemini 3.1 Flash Image, with higher-fidelity generation and faster editing in Gemini and Google AI Studio.
New AI features on Samsung Galaxy S26 including Gemini task handling, enhanced Circle to Search with multi-object recognition, and improved Scam Detection using on-device Gemini model
Redesigned interface with integrated image generation tools (Whisk, ImageFX), flexible asset management, and advanced editing features for video creation in Flow.
Circle to Search now allows searching multiple objects within an image at once, identifying each unique item and providing related information.
Google launched Gemini 3.1 Pro, its upgraded model for more complex reasoning and agentic workflows across the Gemini app, API, Vertex AI, and NotebookLM.
Google added Lyria 3 to the Gemini app, letting users generate short music tracks from text or images with SynthID watermarking built in.
Talking Tours - India Edition uses Gemini to interact with over 200 historical sites, including the Red Fort and Taj Mahal, enabling real-time dialogue about India's heritage.
Expanded Kaggle Game Arena with Werewolf and poker benchmarks to test AI models in social deduction and risk management, with Gemini 3 Pro and Flash leading the leaderboard.
Updated image-to-image generation example to use final gemini-2.5-flash-image model. Upgraded Genkit dependency to version 1.22.
Updated Firebase Security Rules generation to use Gemini 2.5 Pro. Added fixFirestoreRulesFlow tool for auto-correcting Security Rules.
Increased default maximum output tokens from 8,192 to 32,768. Added Gemini 3.0 Flash Preview and 3.0 Pro models. Rolled out Gemini 3.0 Flash to Prototyper workspaces.
Project Genie is an experimental research prototype allowing Google AI Ultra users to create, explore, and remix interactive worlds using text prompts and images.
Gemini navigation features for walking and cycling in Google Maps, including hands-free assistance for route guidance, nearby recommendations, and message sending while exploring or biking.
Chrome gets new Gemini 3 features including auto browse, side panel experience, Connected Apps integrations, and Personal Intelligence capabilities.
Announces Agentic Vision in Gemini 3 Flash — a visual perception capability that enables agent-driven, multi-step visual reasoning and action from images and video (for tasks like scene interpretation and executing visual workflows).
Google AI Pro and Ultra subscribers now receive $10/month and $100/month in Google Cloud credits respectively to build and deploy AI applications, with GDP premium benefits integrated directly into subscriptions.
Google announced Gemini-powered upgrades to AI Mode in Search and to AI Overviews, bringing improved model-generated answers, richer summaries, and enhanced multimodal understanding to Search.
New tools for Google AI Studio and Vertex AI enable animators to integrate generative AI into creative workflows, featuring video-to-video techniques, localized refinement, and model fine-tuning capabilities.
Google introduces Personal Intelligence in AI Mode, allowing AI Pro and Ultra subscribers to connect Gmail and Google Photos for tailored search results.
New features include practice SATs in Gemini, assignment drafting, progress summaries, audio/video recording in Google Classroom, revamped homepage with engagement metrics, and learning standards tagging.
Gemini introduces Personal Intelligence, a beta feature in the U.S. that connects Google apps (Gmail, Photos, YouTube, Search) to provide personalized and helpful answers. Google AI Pro/Ultra subscribers can enable this feature for improved experience and feedback.
Google Trends introduces a new Explore page with Gemini-powered capabilities to identify and compare search trends, offering suggested prompts and a modern interface for deeper insights.
Kaggle launches Community Benchmarks, enabling the AI community to design, run, and share custom evaluations for AI models with tasks ranging from multi-step reasoning to code generation and tool use.
Veo 3.1 updates include native vertical video output (9:16), 1080p/4K upscaling, and improved identity consistency for characters/backgrounds in video generation from images.
New open standard for agentic commerce across platforms, enabling agents and systems to operate together across consumer surfaces, businesses, and payment providers.
Specialized version of Gemma 3 270M model fine-tuned for function calling, designed for edge deployment and customization with support for local execution and API actions.
Introducing Gemini 3 Flash, a fast and cost-effective AI model with frontier intelligence built for speed. Available via Gemini app, Search, and various developer platforms. Offers Pro-grade reasoning at Flash-level efficiency with lower costs.
Introducing GenTabs, a new generative AI web app experience built with Gemini 3, allowing users to create interactive web applications based on open tabs and chat history.
Gemini 2.5 Text-to-Speech models with enhanced expressivity, pacing control, and multi-speaker capabilities for improved voice generation
Pixel Watch 4 introduces AI-powered one-handed gestures (double pinch, wrist turn) and enhanced Smart Replies using the on-device Gemma model for faster, more efficient message responses.
Android XR updates include PC Connect for expanding workspace, travel mode for stable viewing, and Likeness for realistic digital avatars. AI glasses and Project Aura from XREAL are also highlighted.
Gemini 3 Pro is Google's most capable multimodal model, delivering state-of-the-art performance in document, spatial, screen, and video understanding with advanced visual and spatial reasoning capabilities.
Google Ads and Analytics introduce AI advisors to help advertisers with campaign optimization, performance analysis, and data insights. Ads Advisor provides personalized recommendations for campaign management, while Analytics Advisor offers conversational AI for data-driven dec…
Introduces Nano Banana Pro, a new image generation and editing model built on Gemini 3 Pro, enabling accurate visuals with multilingual text rendering and integration across Google products
Introducing Gemini 3, our most intelligent model that combines all of Gemini's capabilities to bring any idea to life.
Google DeepMind's WeatherNext 2 AI model delivers faster, more accurate global weather predictions with 1-hour resolution and 8x speed improvements over previous versions.
AI-based search tool Flight Deals expands to Korea and global markets, enabling smarter flight searches with natural language queries.
New AI-powered features in Google Search enable travel planning with Canvas in AI Mode, global Flight Deals expansion, and agentic AI booking capabilities for restaurants, events, and appointments.
Agentic checkout feature rolls out for Google Shopping, allowing users to track prices and get notified when items are within budget.
New Gemini Live audio updates include adjustable speech speed, language practice tools, interview preparation, storytelling with accents, and fun voice options for enhanced conversational experiences.
New features including AI-powered notification summaries, Wicked: For Good theme packs, and photo remixing in Google Messages
Google Photos introduces Nano Banana AI model for image restyling, AI templates, and expanded Ask Photos features.
Google Labs expanded Opal to over 160 countries, making the no-code AI mini-app builder available globally.
Google Finance introduces AI-powered features including Deep Search, prediction markets data, and live earnings tracking using Gemini models.
Google Maps introduces Gemini-powered features including hands-free navigation, landmark-based directions, proactive traffic alerts, and AI-enhanced Lens for place exploration.
Developers can now use the Google Maps tool in the Gemini API to ground their applications in rich, up-to-date geospatial data, enabling location-aware AI applications.
Veo 3.1 model release with enhanced audio, narrative control, and realism for AI video editing in Flow
Free 12-month Google AI Pro Plan offering advanced Gemini tools to university students in Europe, the Middle East, and Africa
Rolling out Nano Banana, an image editing model in Search via Lens and AI Mode
Introducing Gemini Enterprise, the new front door for AI in the workplace powered by Gemini models, connecting company information to help workplaces collaborate and automate tasks with agents.
The KI-Modus (AI Mode) launches in Germany, Austria, and Switzerland, offering advanced AI-powered search capabilities with logical reasoning and multimodal support (text, image, audio).
Gemini CLI extensions enable developers to customize their command line with integrations from Google partners and open-source tools, enhancing workflow efficiency.
Google Search's AI Mode expands to Spain with enhanced reasoning, multimodal support, and deeper exploration capabilities for complex queries.
A new intelligence for your home powered by Gemini, replacing the Google Assistant and upgrading smart devices like cameras and doorbells.
Gemini 2.5 Flash, a low-latency and cost-efficient model, will soon be available in Vertex AI.
Announcement of Pixel 10, Pixel 10 Pro, Pixel 10 Pro XL, Pixel 10 Pro Fold, Pixel Watch 4, Pixel Buds A Series 2, and Pixelsnap accessories with new Gemini features and design upgrades
Added support for @ symbol to reference folders in chat. Introduced new Python starter template with uv package manager.
Firebase Studio now uses gemini-2.5-flash-image-preview (Nano Banana) model for image generation.
Improved App Prototyping agent features including annotation experience, Select feature refinements, and chat interface improvements.
Gemini is available across devices and in the background, helping with tasks and creativity.
New learning features in Gemini including Guided Learning, Flashcards, and Study Guides.
Enabled Gemini CLI preinstallation. Added ability to add Firebase project directly through App Prototyping agent prompts.
Gemini in Classroom with over 30 no-cost AI tools for educators
Announces new design, Gemini updates, and features across Android ecosystem including Android 16 and Wear OS 6
Firebase Studio now uses Gemini 2.5 for App Prototyping agent and built-in model in Code view chat.
Integration of Gemini features into Google Maps for location inspiration and enhanced navigation tools.
Announced Gemini 2.0, our most capable AI model yet with multimodal output, native image generation, audio output, and tool integration. Experimental version of Gemini 2.0 Flash released for developers. Gemini and Gemini Advanced users can try chat-optimized version. Research pr…
Gemini Pro available via API starting December 13, 2023
Introduction of new Google products including Pixel 4, Pixel Buds, Pixelbook Go, Nest Mini, and Nest Wifi designed for ambient computing and everyday assistance.
Releases per month over the last 12 months. Busiest month: March 2026, 15 releases.
Releases in the last 12 months by category and product.
The most recent Google Gemini update we track is “Guided Vision in Gemini Live for Android”, released on October 1, 2026.
Google Gemini shipped 2 updates in October 2026, including “Guided Vision in Gemini Live for Android”, “Gemini Omni-Powered Asset Studio Tool for YouTube Ad Creation”.
We tracked 94 updates from Google Gemini in the last 12 months, about 8 a month. The busiest month was March 2026, with 15 releases.
Google Gemini publishes release notes on Firebase Studio Release Notes, linked at the top of this page. Every entry here links back to the official page it came from.
Newsletter
This viewscreen turns the calendar into a recurring digest: the sharpest launches, model drops, and product moves delivered in one clear read.
The standout launches, model drops, and platform moves worth tracking.
One tight read instead of bouncing between company blogs and timelines.
Clear summaries you can scan fast and forward to the team when needed.
Monthly dispatch
Get a clean monthly dispatch covering the launches that mattered, not every minor product breadcrumb.
One email each month, tuned for signal over noise.
