Skip to content
ContentLora

    Tip: press / anywhere to search.

    product

    Google Gemini

    Also known as Gemini, Gemini app, Gemini Advanced, Bard, Google Bard

    Gemini is Google's family of multimodal AI models, first introduced on 6 December 2023, and also the name of its AI assistant app, which replaced the Bard chatbot in February 2024.[1][2] The Gemini 4 generation began with Gemini 4 Argon on 30 September 2026.[3] Google calls Gemini 4 Carbon, generally available since 10 November 2026, its strongest model.[4][5] Alphabet's chief executive said in October 2026 that the Gemini app had reached one billion users, a figure the company reports itself.[6]

    Editor reviewedUpdated Wiki35 claims · 100% primaryArtificial intelligence, Big Tech
    Key facts

    Gemini is the name Google uses for two linked things: a family of AI models built by Google DeepMind, and the consumer assistant app that runs on them.[1][2] Google DeepMind was formed in April 2023 by merging Google’s Brain team with DeepMind, and Google introduced Gemini that December.[7][1]

    Model versions

    Google introduced Gemini on 6 December 2023. The first version, Gemini 1.0, came in three sizes: Ultra, Pro and Nano.[1] Google says Gemini was built from the start to be multimodal, handling text, code, audio, images and video. Teams across Google worked on it.[8]

    Google announced these later versions:

    • Gemini 1.5 (15 February 2024): Gemini 1.5 Pro had a standard 128,000-token context window, with up to 1 million tokens in a limited preview.[9]
    • Gemini 2.0 (11 December 2024): Google said it was built for an “agentic era”, with native image and audio output and native tool use.[10] Google said its TPU chips powered all of Gemini 2.0’s training and inference.[11]
    • Gemini 2.5 (25 March 2025): “thinking” models that reason before responding. 2.5 Pro shipped with a 1-million-token context window.[12]
    • Gemini 3 (18 November 2025): Google reported 37.5% on Humanity’s Last Exam without tools and 91.9% on GPQA Diamond for Gemini 3 Pro.[13] It launched alongside Google Antigravity, an agentic development platform.[14]
    • Gemini 4 Argon (30 September 2026): a model aimed at software engineering, enterprise knowledge work and cyber defence.[3] Its output limit is 1 million tokens, up from 64,000.[15] It went first to cyber defence partners through Google’s Fairwind Program.[16] Its introductory API price is $2 per million input tokens and $10 per million output tokens.[17]
    • Gemini 4 Carbon (28 October 2026): Google called it its strongest model, aimed at cybersecurity and long, multi-step tasks, and first offered it to a subset of cyber defenders through the Fairwind Program.[4] Google reported 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym and 71.0% on CWE-bench for patching.[18] It said it took part in the US government’s voluntary pre-release testing process.[19] On 10 November 2026 Carbon became available to all paying API developers at Argon’s introductory price, and began rolling out to Google AI Ultra subscribers in the Gemini app.[5] Google’s price list shows that price doubling to $4 and $20 from 1 December 2026.[20] Google says Argon remains the better choice for most coding and knowledge work, while Carbon, which uses more tokens, is for the hardest tasks.[21]
    • Gemini 4 Hydrogen (12 November 2026): a faster, cheaper Gemini 4 model offered on all tiers of the Gemini API and in the Gemini app and Search, at $1.50 per million input tokens and $7.50 per million output tokens.[22]

    Google also released Gemini 3.8 Flash and a Flash Cyber variant for finding and patching software vulnerabilities in September 2026.[23] In July 2025 an advanced Gemini with Deep Think reached an officially graded gold-medal score at the International Mathematical Olympiad.[24] The benchmark scores above are results reported by Google.[13] For how Gemini compares with rival models over time, see the frontier AI tracker.

    The Gemini app

    Google renamed its Bard chatbot Gemini on 8 February 2024. At the same time it launched a mobile app and Gemini Advanced, which gave access to the Ultra 1.0 model.[2] Gemini Advanced was sold through the Google One AI Premium plan at $19.99 a month.[25] In December 2024 Google added Deep Research, a feature that explores a topic and compiles a report.[26]

    By March 2025 Google said the app was available in over 40 languages and more than 200 countries.[27] Alphabet’s chief executive said it had 950 million monthly active users in July 2026 and one billion users by October 2026; both figures are self-reported by the company.[28][6] In a survey of game developers at GDC 2026, 37% named Google Gemini among the AI tools they used, behind ChatGPT at 74%.[29]

    Gemini in Android, Pixel and Workspace

    In March 2025 Google said it would move mobile Android users from Google Assistant to Gemini and retire the classic Assistant on most mobile devices later that year. It also planned to bring Gemini to tablets, cars, headphones and watches.[30] On Pixel phones, Google says its Tensor G5 chip was the first to run its newest Gemini Nano model fully on the device.[31]

    For businesses, Google included its Gemini AI features in Workspace Business and Enterprise plans in January 2025, without a separate add-on. The features cover Gmail, Docs, Sheets, Meet, Chat and Vids.[32] Consumers had first been offered Gemini in Gmail, Docs, Slides and Sheets through the AI Premium plan.[25]

    Hardware, robotics and regulation

    Google says Gemini runs on its own TPU chips: it said TPUs powered all of Gemini 2.0’s training and inference, and in April 2025 that Gemini 2.5 runs on TPUs.[33][11] Google DeepMind has also built Gemini Robotics on top of Gemini. The first version, released in March 2025, was built on Gemini 2.0.[34]

    Gemini also features in antitrust remedies against Google. On 2 September 2025 a US court barred Google from exclusive contracts to distribute Search, Chrome, Google Assistant and the Gemini app.[35]

    This page covers Google’s Gemini models and app. Robot control models are covered on the Gemini Robotics page, and Google’s AI research group on the Google DeepMind page.

    Google DeepMind

    Google DeepMind is Alphabet's AI research lab, maker of Gemini and AlphaFold. Its origins, leadership, models and 2026 news.

    More in AI

    Questions readers ask

    Is Gemini the same as Bard?

    Yes. Google renamed its Bard chatbot Gemini on 8 February 2024, when it also launched a Gemini mobile app and Gemini Advanced.[2]

    What is the latest Gemini model?

    The current generation is Gemini 4. It began with Gemini 4 Argon, introduced on 30 September 2026 and first rolled out to cyber defence partners, with paid API customers and Google AI Ultra subscribers to follow. Google announced Gemini 4 Carbon, which it calls its strongest model, on 28 October 2026 and made it generally available to paying developers on 10 November. Gemini 4 Hydrogen, a faster and cheaper model for everyday work, followed on 12 November.[3][16][4][5][22]

    How many people use Gemini?

    Alphabet's chief executive said the Gemini app had 950 million monthly active users in July 2026 and had reached one billion users by October 2026. Both figures are self-reported by the company.[28][6]

    Has Gemini replaced Google Assistant?

    On mobile, Google said in March 2025 that it was moving Assistant users to Gemini and would retire the classic Assistant on most mobile devices later that year. It also planned to bring Gemini to tablets, cars, headphones and watches.[30]

    Is Gemini included in Google Workspace?

    Since January 2025, Google has included its Gemini AI features in Workspace Business and Enterprise plans without a separate add-on, across Gmail, Docs, Sheets, Meet, Chat and Vids.[32]

    Sources

    · 35 claims from 10 publishers

    Each numbered claim is a statement we checked against the sources listed with it. Tap a number in the text to see its sources in place.

    1. [1]

      Google introduced Gemini on 6 December 2023; the first version, Gemini 1.0, came in three sizes, Ultra, Pro and Nano. confirmedas of 2023-12-06

    2. [2]

      On 8 February 2024 Google renamed its Bard chatbot Gemini, launched a Gemini mobile app and introduced Gemini Advanced with the Ultra 1.0 model. confirmedas of 2024-02-08

    3. [3]

      On September 30, 2026 Google DeepMind introduced Gemini 4 Argon, a model aimed at software engineering, enterprise knowledge work and cybersecurity defense. confirmedas of 2026-09-30

    4. [4]

      On October 28, 2026 Google announced Gemini 4 Carbon, which it called its strongest model, and made it available first to a subset of cyber defenders through its Fairwind Program. confirmedas of 2026-10-28

    5. [5]

      On November 10, 2026 Google made Gemini 4 Carbon available to all paying developers and enterprises through the Gemini API, at Gemini 4 Argon's introductory price of $2 per million input tokens and $10 per million output tokens, and began rolling it out to Google AI Ultra subscribers in the Gemini app, Antigravity and AI Mode in Search. confirmedas of 2026-11-10

    6. [6]

      Sundar Pichai said in October 2026 that Alphabet was on a half-trillion-dollar revenue run rate, that the Gemini app had reached one billion users, and that YouTube Music and Premium had more than 150 million subscribers including trials. confirmedas of 2026-10-28

    7. [7]

      In April 2023 Google combined the Brain team from Google Research and DeepMind into a single group called Google DeepMind. confirmedas of 2023-04-20

    8. [8]

      Google says Gemini was built from the start to be multimodal, handling text, code, audio, images and video, and was developed by teams across Google. confirmedas of 2023-12-06

    9. [9]

      Google announced Gemini 1.5 on 15 February 2024; Gemini 1.5 Pro had a standard 128,000-token context window, with up to 1 million tokens offered in a limited preview. confirmedas of 2024-02-15

    10. [10]

      On 11 December 2024 Google introduced Gemini 2.0, which it described as built for an "agentic era", with native image and audio output and native tool use. confirmedas of 2024-12-11

    11. [11]

      Google said its TPU chips powered all of Gemini 2.0's training and inference. confirmedas of 2024-12-11

    12. [12]

      Google introduced Gemini 2.5 on 25 March 2025 as a family of "thinking" models that reason before responding; 2.5 Pro shipped with a 1-million-token context window. confirmedas of 2025-03-25

    13. [13]

      Google launched Gemini 3 on 18 November 2025, reporting 37.5% on Humanity's Last Exam without tools and 91.9% on GPQA Diamond for Gemini 3 Pro. confirmedas of 2025-11-18

    14. [14]

      Alongside Gemini 3, Google introduced Google Antigravity, an agentic development platform. confirmedas of 2025-11-18

    15. [15]

      Gemini 4 Argon has a 1 million token output limit, up from 64,000 tokens. confirmedas of 2026-09-30

    16. [16]

      Gemini 4 Argon was first rolled out to cyber defense partners through Google's Fairwind Program, with paid API customers and Google AI Ultra subscribers to follow. confirmedas of 2026-09-30

    17. [17]

      Gemini 4 Argon's introductory API price is $2 per million input tokens and $10 per million output tokens, rising to $4 and $20 after the introductory period; cached inputs are discounted 95 percent. confirmedas of 2026-09-30

    18. [18]

      Google reported that Gemini 4 Carbon scored 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation and 71.0% on CWE-bench for patching. confirmedas of 2026-10-28

    19. [19]

      Google said it was taking part in the US government's voluntary process for pre-release model testing before releasing Gemini 4 Carbon more widely. confirmedas of 2026-10-28

    20. [20]

      Google's Gemini API price list showed Gemini 4 Carbon at $2.00 per million input tokens and $10.00 per million output tokens through November 30, 2026, rising to $4.00 and $20.00 from December 1, 2026. confirmedas of 2026-11-14

    21. [21]

      Google says Gemini 4 Argon remains the better choice for most software engineering and knowledge work, while Gemini 4 Carbon is meant for the hardest tasks and uses more tokens than Argon. confirmedas of 2026-11-10

    22. [22]

      On November 12, 2026 Google introduced Gemini 4 Hydrogen, a faster, lower-cost Gemini 4 model offered on all tiers of the Gemini API and in the Gemini app and Search, priced at $1.50 per million input tokens and $7.50 per million output tokens. confirmedas of 2026-11-12

    23. [23]

      Google also released Gemini 3.8 Flash and a Flash Cyber variant aimed at autonomous vulnerability discovery and automated code patching in September 2026. confirmedas of 2026-10-02

    24. [24]

      In July 2025 an advanced version of Google DeepMind's Gemini with Deep Think achieved an officially graded gold-medal score at the International Mathematical Olympiad, solving five of six problems for 35 of 42 points. confirmedas of 2025-07-21

    25. [25]

      Gemini Advanced launched as part of the Google One AI Premium plan at $19.99 a month, which Google said would also bring Gemini into Gmail, Docs, Slides and Sheets. confirmedas of 2024-02-08

    26. [26]

      Alongside Gemini 2.0, Google launched Deep Research in Gemini Advanced, a feature that explores complex topics and compiles reports for the user. confirmedas of 2024-12-11

    27. [27]

      In March 2025 Google said the Gemini app was available in over 40 languages and more than 200 countries. confirmedas of 2025-03-14

    28. [28]

      As of July 2026 the Gemini app had 950 million monthly active users, according to Alphabet chief executive Sundar Pichai. confirmedas of 2026-07-22

    29. [29]

      Among GDC 2026 respondents, the most used AI tools were ChatGPT (74%), Google Gemini (37%) and Microsoft Copilot (22%), and the most common uses were research or brainstorming (81%), daily tasks such as emails and code assistance (47% each) and prototyping (35%). confirmedas of 2026-01-29

    30. [30]

      In March 2025 Google said it would move mobile Google Assistant users to Gemini, retire the classic Assistant on most mobile devices later that year, and bring Gemini to tablets, cars, headphones and watches. confirmedas of 2025-03-14

    31. [31]

      Google says Tensor G5 was the first chip to run its newest Gemini Nano model, enabling fully on-device generative AI features. confirmedas of 2025-08-20

    32. [32]

      In January 2025 Google included its Gemini AI features in Workspace Business and Enterprise plans without a separate add-on, covering Gmail, Docs, Sheets, Meet, Chat and Vids. confirmedas of 2025-01-15

    33. [33]

      Google said in April 2025 that Gemini 2.5 and AlphaFold run on TPUs. confirmedas of 2025-04-09

    34. [34]

      In March 2025 Google DeepMind introduced Gemini Robotics, a vision-language-action model built on Gemini 2.0, together with Gemini Robotics-ER, an embodied-reasoning model with enhanced spatial understanding. confirmedas of 2025-03-25

    35. [35]

      On September 2, 2025 the court barred Google from exclusive contracts for distributing Search, Chrome, Google Assistant and the Gemini app, and ordered it to share certain search index and user-interaction data and offer search and search text ads syndication to rivals. confirmedas of 2025-09-02

    Revision history (2)
    1. Added Gemini 4 Carbon (announced 28 October 2026, generally available 10 November) and Gemini 4 Hydrogen (12 November); the Latest model fact now lists the Gemini 4 models by role, keeping Argon as the first.
    2. Page created.

    Created Nov 7, 2026. Last reviewed by an editor on Nov 8, 2026. Next scheduled review: Feb 14, 2027.

    Cite this page

    "Google Gemini." ContentLora, updated Nov 14, 2026. https://contentlora.com/wiki/google-gemini

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.