Gemini app

What the app is

The Gemini app is Google's consumer-facing AI assistant. In its basic form it is a chat box: you type a question, and an answer is generated by one of Google's Gemini models. Over time it also became a mobile app, a desktop app, and a subscription product with several paid tiers. It is the company's main way of putting its own models in front of ordinary users, as distinct from offering them to developers through an API.

From Bard to Gemini

The service began as Bard, an experimental conversational AI service announced by Sundar Pichai on 6 February 2023 and initially powered by a lightweight version of LaMDA^[1]. Access opened in the United States and the United Kingdom on 21 March 2023; Google called it an early experiment, and people joined through a waitlist^[2][3]. At Google I/O in May 2023 the product adopted PaLM 2 and expanded to 180 countries^[3]. From 6 December 2023 it used a fine-tuned version of Gemini Pro in English in more than 170 countries and territories^[4].

The Bard name disappeared on 8 February 2024, when the product was renamed Gemini. The same announcement brought a dedicated Android app and Gemini Advanced, a paid tier running Ultra 1.0 at $19.99 a month through the Google One AI Premium plan^[5]. The rename was a consolidation: the assistant and the underlying model family now shared one name.

What it can do

Beyond text chat, the app accumulated distinct capabilities. Gemini Live, previewed at Google I/O in May 2024, launched in August 2024 with the Pixel 9 series^[3]. Deep Research arrived on 11 December 2024 for Gemini Advanced subscribers: an agentic feature that browses the web on the user's behalf and writes up a report with links to its sources^[6]. Later additions included reusable "Skills", introduced on 30 September 2026 to replace the earlier Gems^[9], and connections to more than 40 apps and services, announced on 13 October 2026^[9][10].

Models and subscription tiers

Because the app is the public face of the Gemini model line, new models tend to arrive in it quickly. Pichai said on 18 November 2025 that the app had more than 650 million monthly users; Gemini 3 Pro was made available to everyone in the app that day, and a Gemini Agent feature went to Google AI Ultra subscribers^[7]. On 17 December 2025, Gemini 3 Flash replaced 2.5 Flash as the app's default model for all users^[8].

May 2026 brought a larger batch: Gemini Spark, described as a 24/7 personal agent, Gemini Omni for conversational video creation, Gemini 3.5 Flash, and a Google AI Ultra tier starting at $100 a month^[9]. Gemini 3.6 Flash followed on 21 July 2026^[9]. On 13 October 2026, Gemini 4 Argon became available in the app to Google AI Pro and Ultra subscribers, alongside Google Antigravity in the macOS app and the expanded app connections^[9][10].

Platforms

The app started on the web and on Android. A native Mac app arrived on 15 April 2026, and a Windows app on 10 September 2026^[9].

Why it matters

The app's history tracks a broader shift in how AI assistants are sold. Bard launched as a cautious experiment with a waitlist; within a few years the same product had hundreds of millions of monthly users, a family of named models rotated through it as defaults, and a ladder of consumer subscriptions. For most people it is the first, and sometimes only, place they encounter a frontier model directly, which makes the choice of default model and the design of the paid tiers consequential well beyond Google's own product line.