Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseResearch Paper · For R&DGoogle AI Blog→ original

Google brings Gemini to budget Android Go phones with low memory

Google is bringing Gemini to budget Android Go smartphones, which until now only handled simple voice commands. To make this work, the app was rebuilt so its core features run on devices with limited memory: the Gemini Live voice mode in 45+ languages, camera-based note and form analysis, and image generation with Nano Banana. For developers, the takeaway is simple: AI interfaces need to be tested on weak hardware.

AI-processed from Google AI Blog; edited by Hamidun News
Google brings Gemini to budget Android Go phones with low memory
Source: Google AI Blog. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceGoogle AI Blog ↗
Editor-in-Chief
Key metric #1Gemini Live voice works in more than 45 languages
Key metric #2The camera translates and explains handwritten notes and paper forms
AccessGoogle Play, long-press of the power button, preinstallation with Transsion

Most of the conversation about AI assistants revolves around flagships with tens of gigabytes of memory. But a large part of the world goes online from cheap phones, where every megabyte counts. Google has decided to close this gap: on November 18, the company announced that the Gemini app is now available on eligible Android (Go edition) devices, meaning smartphones with a small amount of RAM.

What has changed

According to Google, until now such phones only offered a basic voice assistant that understood simple commands like "call" or "set an alarm." Now they can run the full Gemini app. The company says it had to rebuild the architecture: the app was reworked so that its main scenario runs smoothly on hardware with less memory.

The announcement contains no technical details. Google does not disclose the minimum amount of memory, or which parts of the computation run on the phone itself and which in the cloud. So the phrase "runs smoothly" remains the company's claim, without independent measurements. The "eligible device" status is also important: launching the app does not mean Gemini will appear on every Go smartphone.

What users get

Google highlights three scenarios.

The first is voice. Instead of typing long queries on an inconvenient on-screen keyboard, you can talk to Gemini Live. For example, you can rehearse a job interview out loud. More than 45 languages are supported. In regions where many people's native language is not English, and typing on a small screen is slow, this is more practical than a text chat.

The second is the camera. You point the phone at a handwritten note or a paper form, and the assistant helps translate the text or work out what it says and what needs to be filled in. For people who deal with documents in a language they don't speak natively, this is an everyday task rather than a feature demo.

The third is images. Through Nano Banana, Google's image generation model, you can make, for example, a flyer for a weekend sale at a bakery. For small businesses without a designer, this is the most obvious example of value.

How to get it, and why it matters for the market

The app can be downloaded from Google Play. On supported phones, Gemini is activated by long-pressing the power button. Google also promises preinstallation on future partner devices, including Transsion, a maker of budget smartphones that are popular in emerging markets. Preinstallation with a partner like this gives Gemini a reach that a single app store could not deliver.

For the industry, we draw several conclusions of our own.

First, competition among assistants is shifting from "whose model is smarter" to "whose assistant is within reach." Whoever gets a foothold on hundreds of millions of inexpensive phones first will win user habits and data on real-world use cases. Second, voice and camera are becoming the main ways to interact with AI where a keyboard is inconvenient. Product teams should factor this into their design. Third, optimizing for low memory is becoming a discipline in its own right, and Google's experience will be useful to anyone building AI clients for the mass market.

Limitations

It is still unknown how the app behaves in real use: how quickly it responds on a weak processor, how it performs on a poor connection, and which features are cut compared with the full version. The list of supported models will also need to be checked separately. Developers and product managers building services for markets with budget phones should test their own apps on this kind of hardware and see how the built-in Gemini performs alongside them. For users, the next step is simple: open Google Play on your Go smartphone and check whether the app is available.

Story

Key Questions Answered

What has changed?

According to Google, until now such phones only offered a basic voice assistant that understood simple commands like "call" or "set an alarm." Now they can run the full Gemini app. The company says it had to rebuild the architecture: the app was reworked so that its main scenario runs smoothly on hardware with less memory. The announcement contains no technical details. Google does not disclose the minimum amount of memory, or which parts of the computation run on the phone itself and which in the cloud.

What users get?

Google highlights three scenarios. The first is voice. Instead of typing long queries on an inconvenient on-screen keyboard, you can talk to Gemini Live. For example, you can rehearse a job interview out loud. More than 45 languages are supported. In regions where many people's native language is not English, and typing on a small screen is slow, this is more practical than a text chat. The second is the camera.

How to get it, and why it matters for the market?

The app can be downloaded from Google Play. On supported phones, Gemini is activated by long-pressing the power button. Google also promises preinstallation on future partner devices, including Transsion, a maker of budget smartphones that are popular in emerging markets. Preinstallation with a partner like this gives Gemini a reach that a single app store could not deliver. For the industry, we draw several conclusions of our own.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…