Gemini

Gemini is Google's generative-artificial-intelligence assistant, chatbot, and consumer-facing AI brand. It originated as Bard, a conversational service announced in February 2023 and initially powered by Google's LaMDA language model. Bard was created as Google's response to the rapid public adoption of OpenAI's ChatGPT and was positioned as a conversational complement to Google Search rather than a conventional search engine. Its early public rollout emphasized experimentation, answer drafting, coding assistance, summarization, and the ability to provide multiple response alternatives while warning users that generated information could be inaccurate or offensive. The underlying technology changed rapidly. Google first moved Bard from LaMDA to PaLM 2 and then incorporated the Gemini model family, which Google announced on December 6, 2023. Gemini was designed as a natively multimodal model family capable of working with combinations of text, code, images, audio, and video. Its variants were organized around different performance and deployment needs: Nano for efficient on-device use, Flash for speed and cost efficiency, Pro for general-purpose capability, and Ultra for demanding workloads. Later generations expanded context windows and introduced stronger reasoning, multimodal generation, tool use, and agentic capabilities. On February 8, 2024, Google unified Bard and its Duet AI productivity branding under the Gemini name. The rebrand also introduced a dedicated Gemini mobile application for Android and an integration into the Google application on iOS. Gemini Advanced, initially associated with Gemini Ultra 1.0 and a Google One AI Premium subscription, became Google's premium consumer offering. Gemini was progressively integrated into Android, Google Workspace, Gmail, Docs, Sheets, Slides, Meet, Messages, Chrome, Photos, Search, Google Cloud, and the Vertex AI developer platform. On compatible Android devices, Gemini also became an alternative or replacement for Google Assistant, while Assistant continued to exist as a separate service during the transition. Gemini serves several overlapping markets. For consumers, it acts as a conversational assistant for writing, studying, planning, research, translation, brainstorming, image creation, document analysis, and voice interaction. For businesses and institutions, it provides AI functions inside Workspace and cloud development products. For software developers, the Gemini API, Google AI Studio, and Vertex AI expose model capabilities for applications, automation, coding, and multimodal workflows. Google has increasingly presented Gemini not merely as a chatbot but as a universal AI interface able to use connected applications, retrieve information, operate across devices, and complete multi-step tasks. The brand has also been shaped by highly visible reliability and safety controversies. Bard's February 2023 announcement was criticized after a product demonstration produced an inaccurate answer about the James Webb Space Telescope. In February 2024, Gemini's image-generation feature produced historically inaccurate and offensive depictions of people, leading Google to suspend image generation of people while it revised the system. Other concerns have involved hallucinations, bias, privacy, copyright, content moderation, contractor working conditions, and the risks of allowing a general-purpose assistant to take actions across connected services. By 2025 and 2026, Google's strategy emphasized Gemini as a broader assistant platform. The company introduced newer model generations, Gemini Live voice interaction, custom Gems, computer-use and agentic functions, Gemini CLI for terminal-based development, Personal Intelligence linked to selected Google services, and Gemini for Home. Google also reported rapid consumer adoption, including more than one billion monthly users for the Gemini app in August 2026. The brand remains active and central to Google's competition in consume…

Gemini's history is rooted in Google's response to the rapid emergence of general-purpose conversational AI. Google had developed large language model research before the public chatbot boom, including LaMDA, announced in 2021 as a conversational language model. LaMDA was not initially released as a general consumer product because Google was concerned about factual reliability, safety, and the consequences of deploying a system capable of producing fluent but incorrect answers. The launch of ChatGPT by OpenAI in November 2022 changed the competitive environment. ChatGPT attracted widespread public attention and raised concerns within Google that conversational AI could challenge the traditional search interface. Google accelerated internal work on a public chatbot, including testing efforts associated with the codename Apprentice Bard. On February 6, 2023, Google announced Bard, an experimental conversational AI service based on LaMDA. Bard was presented as a collaborative tool for drafting, explanation, brainstorming, and information discovery rather than as a replacement for Search. Bard's launch was controversial from the beginning. A February 2023 demonstration included an inaccurate answer concerning the James Webb Space Telescope, and the public presentation was criticized as rushed. Google subsequently opened Bard to a limited number of users in the United States and United Kingdom on March 21, 2023. The early product provided several draft answers, invited user feedback, and displayed warnings that its output could be inaccurate or offensive. Google later moved the service from LaMDA to PaLM 2, improving its mathematical reasoning, coding, multilingual support, and connections to Google products. During 2023, Bard expanded beyond a standalone chatbot. Google added coding assistance, support for more languages and countries, Workspace access, extensions connected to Google services, response verification through Search, and the ability to share conversation threads. Bard's expansion into Europe was delayed while the Irish Data Protection Commission sought additional information about data-protection safeguards. The product eventually became available in the European Union and Brazil with broader language and productivity support. On December 6, 2023, Google introduced Gemini as a new multimodal model family. Gemini was designed to process multiple data types, including text, code, images, audio, and video. Its initial model tiers included Nano, Pro, and Ultra, corresponding broadly to on-device efficiency, general-purpose use, and high-complexity workloads. A version of Gemini Pro was incorporated into Bard, while Bard Advanced was planned as a premium experience using Gemini Ultra. The decisive brand change occurred on February 8, 2024, when Google renamed Bard as Gemini and retired the separate Duet AI branding for Workspace and Cloud. Google launched a Gemini mobile app on Android and integrated Gemini into the Google app on iOS. Gemini Advanced was offered through a paid Google One AI Premium plan. On Android, Gemini could replace Google Assistant as the default assistant experience on supported devices, although the older Assistant remained available during the transition. Gemini's early 2024 development also exposed the difficulty of combining broad safety controls with accurate, context-sensitive generation. Google introduced image generation in Bard shortly before the rebrand, but in February 2024 users reported historically inaccurate and offensive depictions of people. Google acknowledged that the system had overcorrected in some cases and paused image generation of people while it worked on an improved version. The episode became one of the most prominent public controversies associated with the brand and intensified debate about bias, safety evaluation, and model governance. Google then accelerated the technical evolution of Gemini. Gemini 1.5 Pro introduced a much larger context window, enabling analysis of long documents, extensive codebases, lengthy audio, and video. Gemini 1.5 Flash emphasized speed and lower-cost inference. At Google I/O 2024, Google announced broader Gemini integrations across Android, Chrome, Photos, and Workspace, while previewing Gemini Live for voice conversation and Gems for custom assistants. The Gemini 2.0 generation, announced in late 2024, focused on multimodal output, tool use, and agentic behavior. During 2025, Google increasingly described Gemini as a universal assistant rather than merely a chatbot. The company connected it with mobile devices, Workspace, Search, Chrome, coding tools, and developer platforms. Gemini Live expanded, Gemini Advanced gained conversational memory, and Gemini CLI brought an open-source terminal interface to software-development workflows. Google also began shifting more mobile Assistant functions toward Gemini. Google's 2025 and 2026 strategy broadened the product into proactive and personalized assistance. Gemini 3 became available across the Gemini app, AI Studio, and Vertex AI. Personal Intelligence connected Gemini with selected Google services, subject to user control, enabling more tailored recommendations and responses. Google introduced Gemini for Home, expanded connected applications, and announced agentic features such as Gemini Spark. In August 2026, Google stated that the Gemini app had exceeded one billion monthly users. Gemini therefore evolved from a rushed Bard competitor into a central consumer, enterprise, developer, mobile, search, and cloud AI brand within the Google ecosystem.

1. 2026Gemini app becomes more agentic

   Google introduced proactive assistance, daily briefs, Gemini Spark, and new agentic workflows in the Gemini app.

   Source
2. 2026Gemini app surpasses one billion monthly users

   Google reported more than one billion monthly users for the Gemini app.

   Source
3. 2025Mobile Assistant migration toward Gemini

   Google announced a broader transition of mobile assistant functions from Google Assistant to Gemini.

   Source
4. 2025Gemini CLI launches

   Google introduced Gemini CLI as an open-source terminal AI tool for developers and technical workflows.

   Source
5. 2025Gemini 3 launches

   Gemini 3 became available across the Gemini app, Google AI Studio, and Vertex AI.

   Source
6. 2024Bard is renamed Gemini

   Google unified Bard and Duet AI under Gemini and launched dedicated mobile and premium offerings.

   Source
7. 2024Gemini image generation of people is suspended

   Google paused the feature after acknowledging inaccurate and offensive image results involving human subjects.

   Source
8. 2024Gemini 1.5 expands context capacity

   Gemini 1.5 Pro and Flash introduced major improvements in context length, efficiency, and multimodal processing.

   Source
9. 2024Gemini integrations expand at Google I/O

   Google announced Gemini functions across Android, Chrome, Photos, Workspace, voice interaction, and custom assistants.

   Source
10. 2024Gemini 2.0 is introduced

    Google introduced a model generation focused on multimodal output, tool use, speed, and agentic applications.

    Source
11. 2023Bard is announced

    Google announced Bard as an experimental conversational AI service initially based on LaMDA.

    Source
12. 2023Bard opens limited access

    Bard opened to selected users in the United States and United Kingdom through a limited-access waitlist.

    Source
13. 2023Bard adopts PaLM 2

    Google upgraded Bard with PaLM 2 and expanded its language, coding, and productivity capabilities.

    Source
14. 2023Gemini 1.0 is announced

    Google introduced Gemini 1.0 in Nano, Pro, and Ultra versions as a multimodal model family.

    Source
15. 2021Google unveils LaMDA research

    Google presented LaMDA as a conversational language-model research project, establishing an important technical predecessor to Bard and Gemini.

    Source

A broad Google ecosystem AI assistant and model platform combining consumer conversation, multimodal generation, productivity software, mobile assistance, cloud development, coding, search augmentation, and increasingly autonomous task execution.

• Gemini appConsumer AI assistant2024

  The Gemini app is Google's principal consumer interface for conversational and multimodal AI. It supports text interaction, image understanding, document analysis, writing, research, coding, image creation, voice conversation, connected applications, and personalized assistance. It is available through the web and mobile applications and increasingly serves as a front end for Google's newest model generations and agentic features.

• Gemini AdvancedPremium AI assistant2024

  Gemini Advanced is the paid, higher-capability consumer tier associated with Google One AI plans. It was initially introduced with access to Gemini Ultra 1.0 and later moved through newer Pro and advanced model generations. The offering has included larger-context analysis, enhanced reasoning, image and document workflows, conversational memory, priority access, and other premium features.

• Gemini LiveVoice assistant2024

  Gemini Live is a voice-oriented conversational mode designed for natural, interruptible dialogue. It allows users to speak with Gemini in a more continuous manner and supports assistant-style interactions on compatible mobile devices. The feature forms part of Google's effort to replace conventional command-based assistant interactions with richer multimodal and conversational help.

• Gemini for Google WorkspaceEnterprise productivity AI2023

  Gemini for Google Workspace embeds generative AI into products such as Gmail, Docs, Sheets, Slides, Meet, and other productivity services. Capabilities vary by product and plan but include drafting, summarization, rewriting, data analysis, meeting assistance, presentation creation, and enterprise-oriented workflow support. It replaced the separate Duet AI branding used for many earlier Workspace features.

• Gemini APIDeveloper platform2023

  The Gemini API gives developers programmatic access to Google's generative models for text, code, images, audio, video, structured output, and tool-assisted applications. It supports application development through Google's AI tooling and is intended for use cases ranging from chat and content generation to multimodal analysis and autonomous software agents.

• Google AI StudioAI development environment2023

  Google AI Studio is a browser-based environment for experimenting with Gemini models, constructing prompts, testing multimodal inputs, and developing prototypes. It is aimed at developers and technical users who want to evaluate model behavior before integrating Gemini into production applications through APIs or other Google infrastructure.

• Vertex AI GeminiCloud AI platform2023

  Vertex AI provides enterprise and developer access to Gemini models through Google Cloud. It supports model deployment, application development, governance, security controls, evaluation, and integration with other cloud services. The platform positions Gemini as infrastructure for organizations building production AI applications rather than merely as a consumer chatbot.

• Gemini CLIDeveloper tool2025

  Gemini CLI is an open-source terminal-based AI tool designed to assist with software development and computer workflows. It brings Gemini interaction into command-line environments, where users can request coding help, inspect project files, automate tasks, and experiment with agentic development patterns.

• Gemini for HomeSmart-home assistant2025

  Gemini for Home is Google's AI-based voice assistant for home environments and compatible Nest-related experiences. It applies Gemini's conversational and generative capabilities to household questions, routines, device interactions, creative requests, and contextual assistance, extending the Gemini brand from phones and software into connected-home products.

• 2025Gemini as a universal AI assistant

  Global

  Google presented Gemini as a persistent assistant capable of operating across devices, applications, voice interfaces, and computer-use workflows.

  Outcome. The initiative shifted Gemini's positioning toward proactive, multimodal, and agentic assistance.

  Source
• 2024Bard becomes Gemini

  Global

  Google's global rebranding campaign unified Bard, Duet AI, Gemini Advanced, the Gemini mobile app, and related ecosystem integrations under one name.

  Outcome. The campaign established Gemini as Google's principal AI brand across consumer, Workspace, mobile, and cloud products.

  Source
• 2024Google I/O Gemini ecosystem expansion

  Global

  Google used its annual developer conference to present Gemini as an AI layer across Android, Chrome, Photos, Workspace, voice interaction, and custom assistants.

  Outcome. The campaign broadened public understanding of Gemini from a chatbot to an ecosystem-wide assistant and developer platform.

  Source

• 2026Expand Gemini into proactive and personalized assistanceStrategy

  Google sought to make Gemini more useful between explicit prompts by connecting it with user-authorized services and enabling persistent, agentic support.

  What changed. Google introduced Personal Intelligence, Gemini Spark, daily briefs, new connected applications, and additional agentic features.

  Aftermath. Gemini's positioning expanded from a reactive chatbot toward a continuing personal and task-oriented assistant.

  Source
• 2025Transition mobile assistant functions to GeminiStrategy

  Google's traditional Assistant was increasingly limited compared with newer generative and multimodal systems.

  What changed. Google began moving mobile assistant experiences toward Gemini and expanded voice, device, and contextual capabilities.

  Aftermath. Gemini became more deeply embedded in Google's mobile operating-system strategy.

  Source
• 2025Launch Gemini CLIProduct launch

  Google aimed to extend Gemini from browser and mobile interfaces into professional developer workflows.

  What changed. Google launched an open-source command-line tool for coding assistance and terminal-based AI interaction.

  Aftermath. Gemini gained a more direct role in software development and local computer workflows.

  Source
• 2024Unify Bard and Duet AI under GeminiStrategy

  Google had multiple AI brands spanning consumer chat, Workspace, Cloud, and mobile experiences, creating a fragmented product identity.

  What changed. Google renamed Bard as Gemini, retired the Duet AI branding, launched mobile applications, and introduced Gemini Advanced.

  Aftermath. Gemini became the umbrella brand for Google's principal generative-AI experiences.

  Source
• 2024Pause image generation of peopleOther

  Users reported that Gemini generated historically inaccurate or offensive human images and that its safety behavior could overcorrect on demographic prompts.

  What changed. Google temporarily disabled image generation of people and committed to extensive testing and model improvements.

  Aftermath. The incident became a major case study in generative-AI safety, bias, evaluation, and product-release governance.

  Source
• 2024Introduce Gemini 1.5Generation change

  Users and developers needed models capable of analyzing much larger bodies of text, code, audio, and video in a single interaction.

  What changed. Google released Gemini 1.5 Pro and Flash previews with expanded context capacity and improved efficiency.

  Aftermath. Long-context processing became one of Gemini's principal technical differentiators.

  Source
• 2024Introduce Gemini 2.0Generation change

  Google sought to move beyond question-answering toward systems that could use tools, generate multiple modalities, and assist with multi-step tasks.

  What changed. Google introduced Gemini 2.0 with a focus on speed, multimodality, native tool use, and agentic applications.

  Aftermath. The brand's strategic emphasis shifted further toward autonomous and action-oriented AI.

  Source
• 2023Launch Bard as a standalone conversational productProduct launch

  Google needed a public response to the rapid adoption of ChatGPT while remaining cautious about integrating an unreliable generative system directly into the core Search experience.

  What changed. Google released Bard as a separate web application with multiple draft answers, user feedback tools, and explicit warnings about possible inaccuracies.

  Aftermath. The standalone launch enabled rapid experimentation but also exposed reliability, safety, and competitive-positioning problems.

  • Microsoft — Microsoft promoted Bing's AI features and used the launch period to frame the competition as a contest over the future of search.
  • OpenAI — OpenAI's ChatGPT served as the principal market reference point that Google sought to match.
  Source
• 2023Move Bard from LaMDA to PaLM 2Generation change

  Google sought stronger performance in reasoning, coding, mathematics, and multilingual use after Bard's initial LaMDA-based release.

  What changed. Google upgraded Bard's underlying model to PaLM 2 and added expanded coding and language capabilities.

  Aftermath. Bard became more capable and more broadly available before the later Gemini rebrand.

  Source
• 2023Introduce Gemini 1.0Product launch

  Google wanted a unified multimodal model family that could compete across consumer assistants, cloud development, mobile devices, and enterprise applications.

  What changed. Google announced Gemini 1.0 in Nano, Pro, and Ultra variants and integrated Gemini Pro into Bard while preparing Ultra for a premium experience.

  Aftermath. The Gemini name became the foundation for Google's subsequent AI product architecture.

  Source

• 2024Gemini image-generation controversyControversy

  Gemini's image-generation feature produced historically inaccurate and offensive depictions of people, while in other cases it was criticized for refusing or overcorrecting on prompts involving particular demographic groups. Google acknowledged that the system had missed its intended balance and paused image generation of people pending improvements.

  Source
• 2023Bard launch demonstration errorControversy

  Bard's launch presentation included an inaccurate answer about the James Webb Space Telescope. The incident became an early public example of the reliability risks associated with generative AI and contributed to criticism of Google's launch preparation.

  Source
• 2023Concerns over AI training contractorsControversy

  Reporting cited in the public history of Bard raised concerns about working conditions, workload, and compensation among third-party contractors involved in evaluating and training the chatbot. The issue contributed to broader scrutiny of the labor behind generative-AI systems.

  Source

• 2026Gemini expands Personal Intelligence in the United States

  Google expanded an opt-in Personal Intelligence feature that connects Gemini with selected Google services to produce more tailored answers and recommendations.

  Product launch
  Source
• 2026Gemini app becomes more agentic

  Google announced a redesigned Gemini app with proactive assistance, daily briefs, new model capabilities, and Gemini Spark for more persistent task support.

  Product generationProduct launch
  Source
• 2026Google reports more than one billion monthly Gemini app users

  Google reported that the Gemini app had surpassed one billion monthly users and described it as the fastest-growing product in the company's history.

  Other
  Source
• 2026Gemini adds more connected applications and services

  Google announced additional Gemini connections for productivity, travel, dining, entertainment, music, healthcare, and home-service workflows.

  Product launch
  Source
• 2025Google begins transitioning mobile Assistant experiences to Gemini

  Google described the migration of mobile assistant experiences from Google Assistant toward Gemini, with the change intended to provide more capable generative and multimodal assistance.

  Other
  Source
• 2025Gemini adds conversational memory for eligible users

  Gemini Advanced gained a feature allowing it to recall and use information from previous conversations for more personalized interactions.

  Product launch
  Source
• 2025Google introduces Gemini CLI

  Google announced Gemini CLI, an open-source terminal-oriented AI tool intended to assist with software development and computer-based workflows.

  Product launch
  Source
• 2025Gemini becomes a universal AI assistant

  Google outlined a strategy to extend Gemini across devices and services, including computer-use capabilities, personal assistance, and deeper product integration.

  Other
  Source
• 2025Gemini 3 reaches Google products and developer platforms

  Google announced Gemini 3 and made it available across products including the Gemini app, AI Studio, and Vertex AI.

  Product generationProduct launch
  Source
• 2024Bard becomes Gemini

  Google unified Bard and Duet AI under the Gemini brand, launched a Gemini mobile app, and introduced Gemini Advanced with access to Ultra 1.0.

  Product launch
  Source
• 2024Gemini 1.5 Pro introduces a very large context window

  Google previewed Gemini 1.5 Pro with a context window capable of processing very long documents, codebases, audio recordings, and videos.

  Product generation
  Source
• 2024Gemini expands across Android, Chrome, Photos, and Workspace

  Google presented Gemini as a cross-product assistant and announced integrations throughout Android, Chrome, Photos, and Google Workspace.

  Product launch
  Source
• 2024Gemini 2.0 launches for the agentic AI era

  Google introduced Gemini 2.0 with stronger multimodality, tool use, and capabilities intended to support more autonomous AI agents.

  Product generationProduct launch
  Source
• 2023Google announces Bard as a conversational AI service

  Google announced Bard as an experimental conversational AI service powered initially by LaMDA and positioned as a complement to Google Search.

  Product launch
  Source
• 2023Bard opens limited public access

  Bard began limited access in the United States and United Kingdom through a waitlist, with users receiving multiple draft responses and warnings about possible inaccuracies.

  Product launch
  Source
• 2023Bard demonstration error intensifies launch criticism

  A public demonstration included an inaccurate answer about the James Webb Space Telescope, contributing to criticism that Google's launch had been rushed.

  Other
  Source
• 2023Bard adopts PaLM 2 and expands globally

  Google upgraded Bard with PaLM 2, expanded access to additional countries and languages, and added coding, productivity, and Google product integrations.

  Product generationProduct launch
  Source
• 2023Google announces the Gemini model family

  Google introduced Gemini 1.0 in Nano, Pro, and Ultra variants and described the family as natively multimodal.

  Product launchProduct generation
  Source

1. Gemini (chatbot)
2. Introducing Gemini: Google's most capable AI model yet
3. Bard becomes Gemini: Try Ultra 1.0 and a new mobile app today
4. Gemini image generation got it wrong. We'll do better.
5. Google's Gemini 2.0 AI collection
6. Google I/O 2025: Gemini as a universal AI assistant
7. The Gemini app becomes more agentic
8. Google's Gemini app hits 1 billion monthly active users
9. Now you can connect even more of your favorite apps and services to Gemini
10. Google unveils LaMDA research
11. Gemini 1.5 expands context capacity
12. Gemini integrations expand at Google I/O
13. Mobile Assistant migration toward Gemini
14. Gemini CLI launches
15. Gemini 3 launches
16. Gemini在美国扩展个人智能
17. Gemini为符合条件的用户增加对话记忆
18. 谷歌宣布Bard作为对话式AI服务