Google rebrands Bard as Gemini and launches Advanced plan
Google is consolidating its AI brand under Gemini, launching a mobile app and introducing Gemini Advanced, a paid service with access to the Ultra 1.0 model. The move streamlines an offering that until now had divided names and capabilities among Bard, Duet AI and Google One.
Google has retired the Bard brand just a year after introducing its chatbot, replacing it with Gemini. The change comes alongside Gemini Advanced, a subscription with access to the Ultra 1.0 model, and a new mobile app designed to turn the company’s AI into an everyday assistant.
The decision is about more than a simple name change. Google had accumulated several labels for related products: Bard for its chatbot, Gemini for its model family and Duet AI for AI features in Workspace. As of today, Gemini becomes the brand that brings this strategy together in competition with OpenAI’s ChatGPT and Microsoft’s Copilot.
Bard disappears as Gemini becomes the main brand
Bard had already been using Gemini Pro since December, but its name remained tied to the assistant’s initial launch, which was based on a different language model. With the rebrand, the free version is simply called Gemini and retains access to Gemini Pro, the family’s mid-tier model.
Google introduced Gemini in December in three sizes: Nano, designed to run on devices; Pro, intended for general-purpose services; and Ultra, its most capable model. The company says Gemini Ultra outperformed human experts on the MMLU test, a set of academic questions measuring knowledge and reasoning across 57 subjects. In Gemini’s technical report, it scored 90% using a chain-of-thought prompting technique.
That result does not mean the system is better than a person at every task. Benchmarks are useful for comparing models under specific conditions, but they do not capture factors such as the reliability of an answer, the ability to follow ambiguous instructions or the errors that emerge in a long conversation particularly well. Even so, Ultra is the component Google is relying on to compete at the high end of the generative AI assistant market.
Gemini Advanced costs $19.99 a month
Gemini Advanced is the commercial gateway to Ultra 1.0. Google is including it in the new Google One AI Premium plan, which costs $19.99 a month in the United States — €21.99 a month in Spain — and includes two terabytes of Google One storage. The company is offering a two-month free trial at launch.
The service is initially available in English in more than 150 countries and territories. Google is positioning it as an option for tasks that require more reasoning: coding, analyzing long instructions, preparing detailed plans or collaborating on written content.
The subscription also includes generative AI features in Gmail, Docs, Slides, Sheets and Meet. These tools, previously known as Duet AI, will likewise use the Gemini brand. For someone already immersed in Google’s ecosystem, that integration is a meaningful difference from a standalone chatbot: the AI can help inside the apps where users write emails, prepare reports or organize meetings.
The price puts Google in the same league as ChatGPT Plus, which OpenAI sells for $20 a month. The comparison will be unavoidable, although the offerings are not identical: OpenAI has concentrated its product around ChatGPT and its tools, while Google is trying to connect Gemini with Android, Search and Workspace.
An Android app and a gradual arrival on iPhone
Google has also launched the Gemini app for Android. It lets users type, talk to the assistant or attach images, and it can be set as the phone’s primary assistant. In that case, users can invoke it with the power button or a gesture from a corner of the screen, partially replacing the usual way of accessing Google Assistant.
This does not mean Google Assistant is disappearing immediately. When Gemini is set as the primary assistant, Google says some familiar Google Assistant features will also be available. The transition between the two services will be gradual, reflecting a practical limitation of generative AI: it converses and writes better than traditional assistants, but it is not always as predictable when carrying out simple commands.
The Android app is beginning its rollout in English in the United States and will gradually expand to other languages and countries. On iOS, there will not initially be a standalone app: Google will add Gemini to its main iPhone app in the coming weeks.
The challenge will be turning model power into everyday usefulness
Google enters this phase with a distribution advantage that is difficult to match: Android, Chrome, Search and Workspace reach billions of users. But that scale also raises the bar. Gemini will have to show that its answers are useful and consistent when it is integrated into email, personal documents and phones — areas where a mistake can have greater consequences than a wrong answer in a chat.
The initial rollout of Gemini Advanced in English and through a subscription also marks a cautious start. Google is expanding its bet on paid generative AI, but it will still need to add languages, features and availability before the new brand reaches the global scale of the products it aims to unify.
