Keynotesby GlyphGoogle I/O
Google · 4 keynotes · 2026, 2025, 2024, 2023 · mostly Models and Assistants · 25 announcements
Year by year
The top highlight of each keynote20261
Google I/O keynoteGemini 3.5 Flash launched, combining frontier intelligence with agentic action for coding tasksSundar Pichai8 announced1h 51m
20251
Google I/O keynoteGoogle now processes over 480 trillion tokens per month across its productsSundar Pichai8 announced1h 55m
20241
Google KeynoteAI Overviews, the revamped Search results with generative answers, began rolling out to everyone in the USSundar Pichai, Demis Hassabis9 announced1h 53m
20231
Google KeynoteA 2h 5m keynote video from the official Google channel.
Announcements
Gemini 3.5 FlashFrontier intelligence for coding and agentic tasks
Gemini 3.5 ProRolling out next month
Gemini OmniVideo generation from any input; SynthID watermarks
Google Antigravity 2.0Agent-first development platform with CLI and SDK
AI Mode SearchUpgraded with Gemini 3.5 Flash; 1B+ monthly users
Gemini SparkPersonal AI agent; beta for Ultra subscribers
Google PicsImage creation and editing tool launching summer 2026
Android XR Audio GlassesFall 2026 via Gentle Monster, Warby Parker, Samsung
Google BeamAI-powered 3D video communications platform; first devices via HP
Gemini LiveCamera and screen-sharing on Android, rolling out to iOS
Agent Mode (Gemini app)Experimental; automates tasks like apartment hunting
AI Mode (Search)Available now in the US with Gemini 2.5 integration
Gemini 2.5 Pro Deep ThinkExperimental reasoning mode
Veo 3Video generation with native audio, in the Gemini app
Imagen 4Latest image generation model
FlowAI filmmaking tool for cinematic content
AI Overviews in SearchRolling out to everyone in the US that week
Gemini 1.5 Pro 2M context2 million tokens for developers, private preview
Project AstraUniversal AI agent prototype
Imagen 3Image generation model
Veo1080p video generation, select creators via VideoFX
Trillium TPUSixth-generation TPU, 4.7x compute per chip
Gemini 1.5 FlashFast, lower-cost Gemini model
Gemma 2Next generation of the open Gemma models
SynthID for text and videoWatermarking extended to two new modalities
About and year over year
Announcements per year
20268
20258
20249
20230
GAPreviewComingAnnounced
Companies on stage: ElevenLabs, Gentle Monster, Google, Kakao, OpenAI, Samsung, Warby Parker, Anthropic, Automation Anywhere, HP Inc., UiPath, Zillow