Google’s AI history stretches back more than two decades, from machine learning in Search and Google Translate to AlphaGo, the Transformer, BERT, AlphaFold, Bard, and Gemini.
The pace accelerated after Gemini 1.0 arrived in December 2023, with new models and AI products appearing far more frequently.
This timeline tracks key dates from Google’s early AI milestones through Gemini 4 Argon’s October rollout, Nano Banana 2.1, EmbeddingGemma 2, and the Gemini agent for work.
It includes model announcements, research breakthroughs, and product launches in chronological tables.
Last updated: Oct 13, 2026
Gemini Model Release Dates
| Gemini Model | Date | Milestone |
|---|---|---|
| Gemini 4 Argon | Oct 13, 2026 | Broader release for developers, enterprises, and paid subscribers. Google AI Pro and Ultra subscribers gain access in the Gemini app and AI Mode in Search; Antigravity access starts with Ultra subscribers. |
| Gemini 4 Argon Cyber Permissive | Oct 13, 2026 | Generally available defensive cybersecurity variant behind an allowlist. Access requires separate approval through Google or the Fairwind Program. |
| Nano Banana 2.1 | Oct 6, 2026 | Generally available Gemini image generation and editing model based on Gemini 3.6 Flash, with improved visual design, text rendering, and subject consistency. |
| Gemini 4 Argon | Sep 30, 2026 | Initial announcement of the frontier model for software engineering, legal and financial work, and cybersecurity defense. Early access began with trusted cyber defenders through the Fairwind Program. |
| Gemini 3.8 Flash TTS | Sep 22, 2026 | Generally available creative text-to-speech model for custom voice design, directed performances, long-form narration, and multi-speaker dialogue. |
| Gemini 3.8 Flash-Lite TTS | Sep 22, 2026 | Generally available text-to-speech model optimized for cost-efficient, high-volume dubbing, audio production, and voice-agent use. |
| Gemini 3.8 Live | Sep 15, 2026 | Low-latency audio-to-audio model for real-time dialogue, visual context, multilingual conversations, and background tool calls through the Live API. |
| Gemini 3.8 Live Extended Thinking | Sep 15, 2026 | High-reasoning audio-to-audio model with configurable background reasoning and asynchronous tool calls for complex, multi-step voice interactions. |
| Gemini 3.8 Flash | Sep 2, 2026 | Flash model with stronger software engineering, long-running agent, and multi-step reasoning performance at the same introductory API price as Gemini 3.7 Flash. |
| Gemini 3.8 Flash Cyber | Sep 2, 2026 | Cybersecurity model for vulnerability discovery and automated patching, available to trusted defenders through Google’s Fairwind Program. |
| Gemini Omni 1.1 Flash | Aug 27, 2026 | Production-ready video model with scene extension, first-and-last-frame interpolation, faster 360p previews, video references, and output upscaling to 4K. |
| Gemini 3.5 Transcribe | Aug 26, 2026 | Speech-to-text model for real-time and recorded audio with support for more than 85 languages, custom vocabulary, smart cleanup, speaker attribution, and word-level timestamps. |
| Gemini 3.7 Flash | Aug 13, 2026 | New Flash workhorse model for coding, agents, knowledge work, and web development. |
| Gemini 3.6 Flash | Jul 21, 2026 | Flash model focused on stronger coding, knowledge work, multimodal performance, and token efficiency. |
| Gemini 3.5 Flash-Lite | Jul 21, 2026 | Fast 3.5-series model for low-latency, high-volume tasks. |
| Gemini 3.5 Flash Cyber | Jul 21, 2026 | Specialized cybersecurity model paired with Google’s CodeMender security work. |
| Gemini 3.5 Live Translate | Jun 9, 2026 | Audio model for near real-time speech-to-speech translation. |
| Gemini Omni Flash | May 19, 2026 | First Gemini Omni model for multimodal video creation and conversational editing. API public preview followed on Jun 30, 2026. |
| Gemini 3.5 Flash | May 19, 2026 | First model in the Gemini 3.5 family, built for coding and multi-step agent tasks. |
| Gemini 3.1 Flash-Lite | May 7, 2026 | Stable Gemini 3.1 Flash-Lite release for high-volume production use. |
| Gemini 3.1 Flash-Lite | Mar 3, 2026 | Initial preview of the fast, lower-cost Gemini 3.1 model. |
| Gemini 3.1 Pro | Feb 19, 2026 | Gemini 3.1 model for complex reasoning and agent tasks. |
| Gemini 3 Flash | Dec 17, 2025 | Faster Gemini 3 model for everyday use and developer tasks. |
| Gemini 3 Pro | Nov 18, 2025 | First Gemini 3 model released across Google products and developer tools. |
| Gemini 2.5 Flash-Lite | Jul 22, 2025 | Stable low-cost Gemini 2.5 model for latency-sensitive, high-volume use. |
| Gemini 2.5 Flash | Jun 17, 2025 | Stable Gemini 2.5 Flash release for production use. |
| Gemini 2.5 Pro | Jun 17, 2025 | Stable Gemini 2.5 Pro release for production applications. |
| Gemini 2.5 Flash | Apr 17, 2025 | Early Gemini 2.5 Flash release with controllable reasoning. |
| Gemini 2.5 Pro | Mar 25, 2025 | Initial Gemini 2.5 Pro release with stronger reasoning and coding. |
| Gemini 2.0 Flash | Feb 5, 2025 | Updated Gemini 2.0 Flash became available for production through the Gemini API, Google AI Studio, and Vertex AI. |
| Gemini 2.0 Flash | Dec 11, 2024 | Google introduced the first Gemini 2.0 model with multimodal output and native tool use. |
| Gemini 1.5 | Feb 15, 2024 | Gemini 1.5 introduced major long-context improvements, starting with Gemini 1.5 Pro. |
| Gemini 1.0 | Dec 6, 2023 | Original Gemini model family with Ultra, Pro, and Nano sizes. |
Google AI Model & Product Release Timeline
2026 Google AI Releases
| Date | Model or Product | Key Change |
|---|---|---|
| Oct 13, 2026 | Gemini 4 Argon consumer rollout | Google began expanding Argon access to Google AI Pro and Ultra subscribers in the Gemini app and AI Mode in Search, with Google AI Ultra access in Antigravity. |
| Oct 13, 2026 | Gemini 4 Argon developer and enterprise release | Argon became available through the paid Gemini API and Gemini platform under the model ID gemini-4-argon, as well as Gemini for business. Introductory API pricing is $2 per million input tokens and $10 per million output tokens through Nov 30, 2026; standard rates of $4 and $20 apply afterward. |
| Oct 13, 2026 | Gemini 4 Argon Cyber Permissive | Google made the defensive cybersecurity variant generally available behind an allowlist as gemini-4-argon-cyber-permissive. Fairwind approval is required separately from Gemini 3.8 Flash Cyber access. |
| Oct 12, 2026 | Google Health Coach and Fitbit updates | Google detailed improved Coach memory, photo-based meal logging, and deeper Apple Health integration. It also announced Fitbit Edge with Health Coach support; the tracker reaches stores on Oct 27, and coaching requires Google Health Premium. |
| Oct 8, 2026 | Gemini agent for work | Google Cloud introduced a universal work agent at Gemini at Work 2026. It handles knowledge work, media creation, and coding through a single interface, connects to business systems, and selects models for individual tasks under enterprise controls. |
| Oct 8, 2026 | Persistent execution and coworker agents | Google announced cloud execution that continues after a device disconnects, with shared memory across surfaces. Gemini can coordinate temporary sub-agents and persistent coworker agents with dedicated identities and storage. |
| Oct 8, 2026 | Gemini agent in Google Workspace | Google announced inline agent capabilities across Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar, with shared memory, skills, and administrative controls. |
| Oct 8, 2026 | Gemini data and analytics skills | Google introduced skills for PySpark development, editable notebooks, model training, and pipeline repair. Operational reporting skills connect to BigQuery and the Knowledge Catalog for saved queries and department-level cost tracking. |
| Oct 8, 2026 | Gemini industry specializations | Google detailed financial services and legal previews. Finance skills support investment research, credit analysis, and risk modeling; legal integrations inherit matter permissions and ethical walls. Government, healthcare, and retail editions are planned. |
| Oct 8, 2026 | Gemini enterprise governance and cost controls | Google detailed agent identity, authorization, sandboxing, and network controls. Model orchestration, Smart Routing, and real-time project spend caps manage agent costs. |
| Oct 7, 2026 | SynthID Detector public access | Google opened SynthID Detector to everyone globally in English after its earlier release for media professionals. It checks images, video, and audio for SynthID watermarks from Google and participating partners, including OpenAI, NVIDIA, and Kakao. Apple support is planned. |
| Oct 6, 2026 | Nano Banana 2.1 | Google released its updated image generation and conversational editing model as generally available in the Gemini API under the ID gemini-nano-banana-2.1. It improves visual quality, prompt adherence, character consistency, and mask-based editing, with 1K, 2K, and 4K output and new wide and panoramic aspect ratios. Google set Oct 29, 2026 as the API shutdown date for Nano Banana 2 (gemini-3.1-flash-image). |
| Oct 6, 2026 | EmbeddingGemma 2 | Google released a 740-million-parameter open embedding model based on Gemma 4 under the Apache 2.0 license. It maps text, code, images, audio, and video into a shared embedding space for on-device semantic search and retrieval. The model has an 8K-token context window, optional vision and audio encoders, and output vectors that can be reduced from 768 to 128 dimensions. |
| Oct 6, 2026 | Google Flow Music plugin export | Google added VST3 and AU export for custom instruments and effects created with natural-language prompts in Flow Music Spaces. Producers can use the exported plugins inside their digital audio workstations. |
| Oct 1, 2026 | Guided Vision in Gemini Live | Google launched conversational visual assistance on compatible devices running Android 9 or later, in supported Gemini Live regions and languages. You can share your camera for spoken descriptions, follow-up answers, and verbal cues to adjust the framing. Guided Vision supports access through the Gemini app, Android accessibility shortcuts, and TalkBack. |
| Sep 30, 2026 | Gemini 4 Argon | Google announced a new frontier model for long, multi-step coding, enterprise knowledge work, and cybersecurity defense. Google announced support for extended reasoning and long outputs. Initial access began with trusted cyber defenders through the Fairwind Program, ahead of the broader Oct 13 release. |
| Sep 30, 2026 | Skills in Gemini chat | Google began rolling out reusable custom instructions in Gemini chat, including free accounts. You can invoke skills with a slash command, let Gemini apply them automatically, or combine multiple skills. Gems will automatically transition to skills from November 2026 for personal accounts, March 2027 for Workspace business, enterprise, and nonprofit accounts, and June 2027 for education accounts. |
| Sep 24, 2026 | Gemini 3.8 Live with Live Avatar | Google added real-time animated avatars to Gemini 3.8 Live in Gemini Enterprise. The feature combines speech with streaming video, synchronized lip movements, facial expressions, and background tool calls. It supports 97 languages; custom avatar creation requires enterprise allowlisting. |
| Sep 23, 2026 | New Connected Apps in Gemini | Google began rolling out connections to more productivity, creativity, and lifestyle services, including Airtable, Linear, Adobe, Webflow, Peloton, and SeatGeek. You can connect supported apps in Gemini settings or call them from a chat. |
| Sep 23, 2026 | Gemini 3.8 Flash TTS Gemini 3.8 Flash-Lite TTS | Google announced two text-to-speech models for expressive voice design, line-by-line performance direction, and high-volume audio production. The Gemini API lists both as generally available from Sep 22; Google AI Studio access began rolling out with the Sep 23 announcement, alongside Flash TTS in Gemini Notebook and Flash-Lite TTS in Google Vids. |
| Sep 19, 2026 | Google CC | Google Labs introduced new Google CC, a new experimental agent built for families, helping households spend less time on logistics and more time together. |
| Sep 15, 2026 | Gemini 3.8 Live Gemini 3.8 Live Extended Thinking | Google introduced two real-time audio models for voice agents and live dialogue. Gemini 3.8 Live focuses on low-latency conversation, visual context, multilingual interaction, and background tool calls. Gemini 3.8 Live Extended Thinking adds configurable background reasoning and asynchronous tool use for complex multi-step tasks. Both began rolling out through the Gemini API and Google AI Studio, with additional availability across Search Live, Gemini Live, Workspace, and Gemini Enterprise. |
| Sep 10, 2026 | Gemini app for Windows | Google released a native Gemini desktop app for Windows with an Alt + Space shortcut, access to connected Google apps, task assistance, and image and video creation from the desktop. |
| Sep 4, 2026 | Lyria 3.5 in Gemini and the Gemini API | Google expanded Lyria 3.5 music generation to the Gemini app and Gemini API after its earlier Flow Music release, with richer vocals, more detailed arrangements, and selectable track lengths. |
| Sep 3, 2026 | WeatherNext 3 | Google DeepMind introduced its flagship weather forecasting AI model with real-time satellite data, hourly refreshes, higher-resolution forecasts, more precise precipitation prediction, and clean-energy variables. WeatherNext 3 is integrated across Search, Gemini, Maps, Google Maps Platform, and Google Cloud. |
| Sep 3, 2026 | Gemini Audio in Gmail, Docs, and Keep | Google launched real-time voice features across Gmail, Docs, and Keep for conversational inbox search, spoken drafting, and voice capture in eligible Google AI plans. |
| Sep 2, 2026 | Gemini 3.8 Flash and Gemini 3.8 Flash Cyber | Google introduced Gemini 3.8 Flash for long-horizon software engineering, autonomous agents, and multi-step reasoning, plus a Flash Cyber model for vulnerability discovery and automated patching. Gemini 3.8 Flash launched at an introductory API price of $0.75 per million input tokens and $3.75 per million output tokens through Dec 31, 2026; Google says those rates will rise to $1.50 and $7.50 on Jan 1, 2027. Flash Cyber access is limited to trusted defenders through the Fairwind Program. |
| Sep 1, 2026 | Agentic video understanding | Google launched agentic video understanding for Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite through the Gemini API in Google AI Studio and Gemini Enterprise Agent Platform. Gemini can navigate a video timeline and request transcripts, frames, or audio as needed, with up to 88% fewer tokens, up to 66% lower analysis costs, and up to 7% higher accuracy in Google’s benchmarks. |
| Sep 1, 2026 | Google Pics | Google Workspace made its Nano Banana-powered image creation and editing tool generally available. Pics can generate and refine images, edit selected objects and in-image text, crop and upscale images, keep version history, and support collaboration, with direct editing in Docs and Slides and Drive integration rolling out in the coming weeks. |
| Aug 27, 2026 | Gemini Omni 1.1 Flash | Google made its updated video model production-ready with scene extension up to 40 seconds, first-and-last-frame interpolation, faster 360p previews, video references, and 1080p or 4K output. |
| Aug 27, 2026 | Expert Intelligence in Gemini Notebook | Google began supporting eligible ebooks purchased through Google Play Books in Gemini Notebook, with responses and generated materials grounded in the book. |
| Aug 26, 2026 | Gemini 3.5 Transcribe | Google released a speech-to-text model for live and recorded audio in public preview, with support for more than 85 languages, custom vocabulary, smart cleanup, and detailed timestamps. |
| Aug 25, 2026 | Gemini Enterprise for Legal and Financial Services | Google Cloud introduced preview industry solutions with governed agents, domain-specific skills, secure data connectors, and partner integrations for legal and financial work. |
| Aug 20, 2026 | Google Antigravity enterprise subscriptions and IDE extensions | Google included Antigravity in eligible Gemini Enterprise subscriptions with pooled quotas, budget caps, and centralized security controls. New extensions support VS Code, with Visual Studio, JetBrains, and Zed in preview, alongside the desktop app and CLI. |
| Aug 13, 2026 | Gemini 3.7 Flash | Google released a new Flash workhorse model for coding, agents, knowledge work, and web development. |
| Jul 28, 2026 | Managed Agents in Gemini API | Managed Agents moved to Gemini 3.6 Flash by default and gained environment hooks, budget controls, scheduled triggers, and free-tier access. |
| Jul 21, 2026 | Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber | Google expanded the Flash family for production agents, low-latency tasks, and specialized cybersecurity work. |
| Jun 30, 2026 | Nano Banana 2 Lite and Gemini Omni Flash API preview | Google released a faster Gemini image model and opened Gemini Omni Flash to developers and enterprises in public preview. |
| Jun 24, 2026 | Computer use in Gemini 3.5 Flash | Gemini 3.5 Flash gained built-in computer use for agents that interact with desktop, mobile, and browser environments. |
| Jun 10, 2026 | DiffusionGemma | Google introduced an experimental open model focused on faster local inference. |
| Jun 9, 2026 | Gemini 3.5 Live Translate | Google released an audio model for near real-time speech-to-speech translation across more than 70 languages. |
| Jun 4, 2026 | Dreambeans | Google Labs introduced a mobile AI experiment that creates personalized daily stories from connected Google apps. |
| Jun 4, 2026 | Gemma 4 12B | Google announced a multimodal Gemma model designed for capable local use on laptops. |
| May 19, 2026 | Gemini 3.5 Flash and Gemini Omni | Google I/O 2026 introduced a new Gemini generation for agentic work and a multimodal family that starts with video generation and editing. |
| May 19, 2026 | Gemini Spark | Google introduced a personal AI agent designed to work across Google apps and supported partner services. |
| May 19, 2026 | Google Antigravity 2.0 | Google expanded its agent-first development platform with a standalone desktop app, multi-agent orchestration, scheduled tasks, and product integrations. |
| May 19, 2026 | Universal Cart | Google announced an intelligent shopping cart that works across participating Google surfaces and merchants. |
| May 19, 2026 | AI information agents in Search | Google introduced Search agents that can monitor ongoing tasks, handle follow-ups, and provide updates. |
| May 19, 2026 | Intelligent eyewear | Google previewed Gemini-powered eyewear with audio-first and display-based experiences. |
| May 12, 2026 | Gemini Intelligence on Android | Google introduced more proactive Gemini features across Android devices and multi-step actions across apps. |
| May 12, 2026 | Googlebook | Google introduced a new laptop category designed around Gemini Intelligence and Android integration. |
| May 5, 2026 | Gemini API File Search | File Search gained multimodal support, custom metadata, and page-level citations. |
| Apr 22, 2026 | Gemini Enterprise Agent Platform | Google launched an enterprise platform for building, managing, and governing AI agents. |
| Apr 22, 2026 | Gemini Embedding 2 | Google made its multimodal embedding model generally available. |
| Apr 21, 2026 | Deep Research Max | Google released a more capable research agent for longer and more complex research tasks. |
| Apr 15, 2026 | Gemini on Mac | Gemini became available as a desktop assistant for macOS. |
| Apr 14, 2026 | Skills in Chrome | Chrome gained reusable prompt-based AI actions for repeated browser tasks. |
| Mar 31, 2026 | Veo 3.1 Lite | Google introduced a lower-cost Veo model for developer video generation. |
| Mar 25, 2026 | Lyria 3 Pro | Google released a music generation model for longer tracks and greater composition control. |
| Mar 11, 2026 | Gemini Embedding 2 preview | Google introduced a multimodal embedding model for text, images, video, and audio. |
| Feb 26, 2026 | Nano Banana 2 | Google introduced Gemini 3.1 Flash Image for faster image generation and editing. |
| Jan 14, 2026 | Personal Intelligence in Gemini | Gemini gained a beta personalization feature that can use context from connected Google apps. |
2025 Google AI Releases
| Date | Model or Product | Key Change |
|---|---|---|
| Dec 17, 2025 | Google Workspace Studio | Google introduced a Workspace product for creating work automation agents without code. |
| Nov 18, 2025 | Gemini 3 and Google Antigravity | Google launched the Gemini 3 generation and introduced an agent-first development platform. |
| Nov 6, 2025 | Gemini API File Search Tool | Google added managed retrieval directly to the Gemini API. |
| Nov 6, 2025 | Opal global expansion | Google expanded its AI mini-app builder to more than 160 countries. |
| Oct 28, 2025 | Pomelli | Google released an experimental AI marketing tool for creating on-brand content. |
| Oct 15, 2025 | Veo 3.1 | Google updated its video generation family with richer audio and more control. |
| Sep 25, 2025 | Search Live | Google Search added real-time voice and camera-based AI answers. |
| Sep 9, 2025 | Gemini Canvas visual editing | Gemini Canvas gained click-to-edit visual changes for generated web apps. |
| Jul 25, 2025 | Opal | Google Labs released a visual tool for building and sharing AI mini-apps. |
| Jul 14, 2025 | Gemini Embedding | Google’s Gemini Embedding text model became generally available for developers. |
| Jun 25, 2025 | Gemini CLI | Google introduced an open-source terminal agent for using Gemini from the command line. |
| Jun 6, 2025 | Scheduled actions in Gemini | The Gemini app gained scheduled tasks and recurring updates. |
| May 31, 2025 | Gemini Live camera and screen sharing | Gemini Live expanded camera and screen-sharing access across Android and iOS. |
| May 27, 2025 | Gemini 2.5 Flash native audio | Google expanded Gemini 2.5 Flash with native audio output through the Live API. |
| May 27, 2025 | AI Overviews expansion | AI Overviews expanded to more countries, territories, and languages. |
| May 20, 2025 | Veo 3, Imagen 4, Flow, and Gemma 3n | Google I/O 2025 expanded its AI portfolio across video, image generation, filmmaking, and open models. |
| May 14, 2025 | Gemini Advanced GitHub connection | Gemini Advanced gained GitHub repository context for coding tasks. |
| Apr 29, 2025 | NotebookLM multilingual Audio Overviews | NotebookLM expanded Audio Overviews to more than 50 languages. |
| Apr 16, 2025 | Google AI Studio Starter Apps | Google AI Studio added editable starter apps for model-powered development. |
| Apr 15, 2025 | Veo 2 in Gemini Advanced | Google brought Veo 2 video generation into Gemini Advanced. |
| Apr 9, 2025 | Deep Research with Gemini 2.5 Pro | Gemini Advanced gained Deep Research powered by Gemini 2.5 Pro Experimental. |
2024 Google AI Releases
| Date | Model or Product | Key Change |
|---|---|---|
| Dec 13, 2024 | Google Agentspace | Google Cloud introduced an enterprise AI search and agent product. |
| Dec 4, 2024 | Genie 2 | Google DeepMind introduced a foundation world model for generating interactive environments. |
| Oct 9, 2024 | Imagen 3 | Google expanded access to its higher-quality image generation model. |
| Sep 2024 | Ask Photos | Google Photos introduced a conversational way to search photos and videos. |
| Sep 2024 | DataGemma | Google introduced research that connects language models with real-world statistical data. |
| Aug 2024 | Gemini overlay on Android | Gemini expanded as an Android assistant overlay for in-context help. |
| Jul 31, 2024 | Gemma 2 2B, ShieldGemma, and Gemma Scope | Google expanded the Gemma family with smaller, safety, and interpretability resources. |
| Apr 2024 | CodeGemma | Google released Gemma models tuned for code completion and generation. |
| Mar 2024 | TacticAI and SIMA | Google DeepMind published AI systems for football tactics and generalist game agents. |
| Feb 2024 | Genie | Google DeepMind introduced a foundation world model for generated interactive environments. |
| Jan 2024 | MobileDiffusion | Google published research on faster on-device text-to-image generation. |
Earlier Google AI Milestones: 2001-2023
| Date | Google AI Milestone |
|---|---|
| 2001 | Machine learning helped Google Search suggest better spellings for queries. |
| 2006 | Google Translate launched and brought machine translation to a large public audience. |
| 2015 | Google released TensorFlow as an open-source machine learning framework. |
| 2015 | Google launched RankBrain, its first deep learning system deployed in Search. |
| 2016 | AlphaGo defeated Lee Sedol in Go, one of DeepMind’s most visible early AI milestones. |
| 2016 | Google’s Tensor Processing Units helped accelerate large-scale machine learning training and inference. |
| 2017 | Google researchers introduced the Transformer architecture, which became a foundation for modern language models. |
| 2019 | BERT improved how Google Search understands language, context, and query meaning. |
| 2020 | AlphaFold reached a major breakthrough in protein structure prediction. |
| May 18, 2021 | Google introduced LaMDA, a Transformer-based language model designed for open-ended dialogue. |
| 2023 | Bard launched as Google’s consumer generative AI assistant. |
| May 10, 2023 | Google introduced PaLM 2 and expanded generative AI across Bard, Workspace, Cloud, and other products. |
| May 10, 2023 | Google introduced Search Generative Experience in Search Labs. |
| Dec 6, 2023 | Google introduced Gemini 1.0 with Ultra, Pro, and Nano model sizes. |
Major Google AI Models, Platforms & Products
| Name | Type | Primary Role |
|---|---|---|
| Gemini | Model family / AI assistant | Google’s main multimodal model family and consumer AI assistant. Gemini 4 Argon focuses on complex coding, professional knowledge work, and cybersecurity defense. |
| Gemini agent | Enterprise AI agent | Google Cloud work agent for knowledge tasks, media creation, and coding across connected business systems. |
| SynthID Detector | AI media verification | Checks supported images, video, and audio for SynthID watermarks from Google and participating partners. |
| Gemini Audio | Model family | Google’s audio and speech model family, including Gemini 3.8 Flash TTS and Flash-Lite TTS for speech generation, Gemini 3.8 Live models for real-time dialogue, and Gemini 3.5 Transcribe and 3.5 Live Translate. |
| Gemma | Open model family | Google’s open model family for local, research, and production use. |
| EmbeddingGemma | Open embedding model family | Models for local semantic search and retrieval. EmbeddingGemma 2 supports text, code, images, audio, and video in a shared embedding space. |
| Gemini Omni | Model family | Multimodal model family that starts with video generation and conversational editing. |
| Veo | Model family | Google’s video generation model family. |
| Imagen / Nano Banana | Image model family | Google image generation and editing models used across consumer and developer products. Nano Banana 2.1 is based on Gemini 3.6 Flash and improves visual design, mask-based editing, and subject consistency. |
| Google Pics | Workspace product | Nano Banana-powered Google Workspace tool for AI image creation, precise object editing, in-image text changes, and collaboration. |
| Google DeepMind | Research organization | Google’s central AI research organization behind Gemini, AlphaGo, AlphaFold, Genie, and other major research programs. |
| Google AI Studio | Developer platform | Developer environment for testing Gemini models and building with the Gemini API, including agentic video analysis. |
| Vertex AI | Cloud AI platform | Google Cloud platform for building and deploying AI applications and models. |
| Gemini Enterprise | Enterprise agent platform | Google Cloud’s governed environment for creating, deploying, and using AI agents, including purpose-built editions for specific industries. |
| Gemini CLI | Open-source coding agent | Open-source terminal AI agent for coding, research, and command-line tasks. |
| Google Antigravity | Developer platform | Agent-first development platform for building and coordinating software agents. |
| Gemini Skills | Customization feature | Reusable instructions that Gemini can apply automatically or you can invoke and combine in a chat. Skills will replace Gems through a phased account migration. |
| Gemini Spark | AI agent | Personal AI agent for proactive tasks across Google apps and supported services. |
| NotebookLM | Research product | Source-grounded research and note-taking product built around materials you provide. |
Related AI Timeline Resources
- AI Model Release Calendar: Latest AI Model Releases by Date
- OpenAI & ChatGPT Timeline: GPT Release Dates to GPT-6 Astra
- Claude Timeline: Model Release Dates and Version History
- Qwen Timeline: Model & Product Release History
- DeepSeek Timeline: Model Release Dates and Key Milestones
- Meta AI Timeline: FAIR to Llama 4 and Muse AI
- SpaceXAI Timeline: Grok Release History and Product Milestones
- Mistral AI Timeline: Model and Product Release History
- Kimi Model Timeline: Kimi K3 Release Dates and Version History
- Codex Timeline: Release Dates and Major Updates (2026)
- Claude Code Timeline: Release Date and Major Updates (2026)
- Apple Intelligence & Siri AI Timeline: Release Dates & Updates
- Midjourney Timeline: Version Release Dates and Major Updates
- Manus AI Timeline: Release Dates and Major Updates
More Google AI Tools & Resources
- Code Wiki: Google’s AI-Powered Solution for Code Understanding.
- NotebookLM: AI Research Assistant for Students and Professionals.
- Pomelli: Google’s Free AI Tool for SMB Marketing Campaigns
- Google Stitch: Free AI Vibe Design Tool with Agent and Web Export
- Google Antigravity 2: Agent-First Dev Platform with Gemini 3.5.
- Dreambeans: Daily AI Stories from Google Apps
For current API cost comparisons across Google, OpenAI, Anthropic, and other providers, see the AI LLM API pricing guide.
FAQs
When was Nano Banana 2.1 released?
Nano Banana 2.1 became generally available in the Gemini API on Oct 6, 2026 as gemini-nano-banana-2.1. It updates Nano Banana 2 with better image quality, text rendering, and consistency across repeated edits. The previous API model, gemini-3.1-flash-image, is scheduled to shut down on Oct 29, 2026.
What is EmbeddingGemma 2?
EmbeddingGemma 2 is Google’s open multimodal embedding model, released on Oct 6, 2026. It converts text, code, images, audio, and video into numerical vectors for local search and retrieval. The full model has 740 million parameters and uses an Apache 2.0 license. Developers can use it for offline media search, code retrieval, and retrieval-augmented generation.
When was Gemini 4 Argon announced, and who can use it?
Google announced Gemini 4 Argon on Sep 30, 2026 and expanded availability on Oct 13. Google AI Pro and Ultra subscribers can access it in the Gemini app and AI Mode in Search. Developers and enterprises can use the paid API, while Antigravity access starts with Ultra subscribers. The Cyber Permissive variant requires separate allowlist approval.
Can Gemini 4 Argon produce 1 million output tokens?
Argon has a shared 1,048,576-token window for input, thinking, and output. Long decode can extend generation across continuation requests within that window. A single streaming request returns up to 65,536 tokens, and a non-streaming GenerateContent request returns up to 32,768. The available output therefore depends on the input length and tokens used for thinking.
Are Gemini Skills replacing Gems?
Yes. Gems will automatically become skills from November 2026 for personal Google accounts, March 2027 for Workspace business, enterprise, and nonprofit accounts, and June 2027 for education accounts. Skills can run within an existing chat and can be combined for a task. They do not yet support every Gem feature, including Canvas and Deep Research.
Why can the same Gemini model have more than one release date?
Google can announce a model before it reaches general availability. Experimental access, preview releases, API availability, and general availability can therefore have different dates. This timeline keeps those milestones separate when the distinction changes when or how the model can be used.
When was Google Gemini first released?
Google introduced Gemini 1.0 on Dec 6, 2023. The original family included Gemini Ultra, Gemini Pro, and Gemini Nano.
When was Gemini 2.0 released, and when did Gemini 2.0 Flash become generally available?
Google introduced Gemini 2.0 on Dec 11, 2024 with an experimental release of Gemini 2.0 Flash. Gemini 2.0 Flash reached general availability on Feb 5, 2025 through the Gemini API in Google AI Studio and Vertex AI.
When did Bard become Gemini?
Google renamed Bard to Gemini in Feb 2024 and expanded the Gemini app experience across mobile platforms.
When did Google start using AI in Search?
Google used machine learning in Search long before generative AI. Spelling correction appeared in 2001, RankBrain launched in 2015, BERT reached Search in 2019, Search Generative Experience arrived in 2023, and AI Overviews followed in 2024.
