Google Gemini AI explained: Cost, features, availability, and controversies
From Gemini 1.5 Flash to Pro, Gemini Advanced to Gemma and Vertix, here's every Gemini AI feature and version you need to know about.
Google Gemini is a generative AI chatbot and Large Language Model (LLM). Formerly Google Bard, the Gemini AI can generate stories from prompts, give you answers to coding problems, search an uploaded image for information, or give you contextual information based on your location or data from other Google apps.
At Google I/O 2024, the company spent a looooong time breaking down its future AI plans, including several new Gemini versions and upgrades. At this point, it's not just Gemini; it's Gemini Nano, Flash, Pro, Advanced, Live, the Android Gemini app, and several other versions.
That's not to mention other Google AI projects related to Gemini AI, like Google AI Studio, Gemma, and Project Astra. It's hard to keep them all straight! That's where we come in.
Below, we'll break down how the different versions of Gemini work, where you can access them, whether you should pay for Gemini Advanced, and everything else you need to know.
How Google Gemini works
Google Gemini, like Google Search, responds to your questions by pulling from internet sources to answer them. The difference is that Gemini tries to format its own conversational responses instead of highlighting third-party sources.
Using Google's Language Model for Dialogue Applications (LaMDA) and its open-source "Transformer" machine-learning model, Gemini "reads" trillions of words from every publicly available source — then uses it to mimic human responses while rewriting information from the internet.
When you see a Gemini AI Overview in your Search results, it summarizes data from a few high-ranking sources, with the links visible under the "Show More" button. But the "AI" can't actually determine if these sources are correct; that's why you'll occasionally get incorrect AI responses based on satirical websites or trolling Reddit comments.
Be an expert in 5 minutes
Get the latest news from Android Central, your trusted companion in the world of Android
The quality of the Gemini AI response will also vary based on which version of Gemini you're using.
Where the name "Gemini" came from
Google called its AI "Bard" before replacing the name with "Gemini." It's fair to wonder why!
Gemini's Latin meaning is "twins," and it's famously a Zodiac constellation with pair stars. It's also the name of NASA's original moonshot program before the famous Apollo missions, with Gemini spacecraft having enough room for two.
According to a Google blog post, "How Google’s AI model Gemini got its name," the Gemini AI name stemmed from how two Google departments (DeepMind and the Google Research Brain team) joined forces for the project, and because "the monumental effort of training LLMs resonated with the spirit of launching rockets."
Bard isn't mentioned in the post, so on a less thematic level, the name Gemini probably sounds more futuristic and marketable than Bard.
Google Gemini versions and spin-offs
Below, we'll briefly summarize the different Google Gemini versions you need to know about:
Gemini Nano: Designed for on-device smartphone applications that work using a neural processing unit (NPU) without needing a network connection. Available on phones like the Pixel 8 Pro, Gemini Nano powered features in the December 2023 feature drop like smart replies in Whatsapp or a summary tool for the Recorder app.
Gemini 1.5 Pro: The standard Gemini experience, with a 1 million token context window (meaning it can interpret a lot of text, code, audio, or video simultaneously for a single question). Developers can currently try a Pro version with 2 million token context. Gemini Pro is what companies like Samsung use to add Gemini AI into its apps like Samsung Notes, Voice Recorder, and Keyboard.
Gemini 1.5 Flash: This version uses the same 1 million standard as 1.5 Pro, but it's "optimized for narrower or high-frequency tasks where the speed of the model’s response time matters the most." Like Pro, it's available through the Gemini API to developers.
Gemini Advanced: Gemini Advanced is Google's subscription for everyday Google One users to access its AI tools. It uses Gemini 1.5 Pro and its 1 million token context window, so it's not more "advanced" for smarts. However, it does give you specific tools like Google Docs and Sheets summaries, Gemini Live conversations, and custom Gemini versions called "Gems."
Gemini Live: Coming to Gemini Advanced later this year, Gemini Live is simply a voice-first version of Gemini Pro that lets you talk to the AI chatbot like a voice assistant on your Android phone. You'll be able to choose between 10 different voices and "interrupt mid-response with clarifying questions" if you want.
Gems: Another future feature for Gemini Advanced subscribers, Gems are customized versions of Gemini you create to your own specifications. You can give parameters like "you're my running coach, give me a daily running plan and be positive, upbeat and motivating" and then speak to a "Gem" with that tone and underlying goal.
Gemma: Google's open AI spin-off of Gemini. Different Gemma builds "share technical and infrastructure components" with Gemini, except developers can use the code to build their own AI projects. Specific variants like CodeGemma (for coders) and RecurrentGemma (for researchers) are more specialized, while Gemma 2 (out June 2024) is a more general smarts upgrade.
Vertix AI: An AI development platform that includes both Gemini 1.5 Flash and 1.5 Pro, as well as 150+ other AI "foundation models."
Imagen: Announced at Google I/O 2024, Imagen 3 is a text-to-image model that Google says will come to Gemini's web and app versions in the future. It can produce photorealistic art based on text prompts, like the image below. At the moment, you can join a waitlist to try it out.
Veo: Google's text-to-video model that "generates high-quality, 1080p resolution videos that can go beyond a minute, in a wide range of cinematic and visual styles." It will come to the VideoFX tool and YouTube Shorts, and is built off of several AI models (including Gemini).
Project Astra: Gemini 1.5 Pro can handle multimedia, but is mainly designed to interpret one thing at a time. Project Astra is a multimodal AI that can answer a contextual question like "What am I looking at?" based on visual data. This tech could help power future Google AR glasses. The DeepMind team says Astra "builds on our Gemini models," and some features could come to the Gemini AI app later in 2024.
Google AI Studio: This website is where you can go to start working on Gemini projects, with the Gemini API available and tools like a code "Cookbook" and prompts available to help you get started.
Where is Google Gemini available?
The standard version of the Google Gemini AI is "currently available in more than 40 languages and over 230 countries and territories," according to a Google Support page. After Google brought Gemini to Canada after a legal delay, it has very few territories left where it doesn't support the Gemini AI.
The Gemini mobile app is a bit more limited. The Support availability page currently shows 19 languages and "more than 150 countries."
In either case, you need a legitimate Google account to access it, as mentioned above. You must be 18 or older in the European Economic Area (EEA), Canada, Switzerland, or the UK; in other areas, you must be at least 13 — or 14–17 in specific countries — with an unrestricted account.
How to access Google Gemini
Go to gemini.google.com/app to access the generative AI on any device, across all available languages and countries (more on that later).
If you're Google account is controlled under Family Link, Google Workspace for Education, or a Workspace where your administrator hasn't enabled Gemini, then you won't be able to access the tool. You need a valid Google account for it to work.
Although some Gemini tools are Pixel 8-exclusive, you can download the Google Gemini Android app to access the AI directly on most of the best Android phones running Android 10 or later. Unfortunately, we've discovered that certain Android phones don't let you install Gemini, such as the Motorola Razr. Google says that "we are working on it and aim to have it available soon.
On iPhones and iPads, you can access the Gemini AI through the standard Google app.
Chromebook Plus models will receive Gemini features in a June 2024 update, including Help Me Write, Generative AI Wallpaper and Video Call Backgrounds, and Magic Editor on Google Photos.
Lastly, you should know how to control Gemini Extensions to other Google apps like Maps, Hotels, Flights, YouTube, YouTube Music, and Workspace. This Gemini Extensions management page lets you toggle specific apps on and off, as needed.
Gemini Advanced AI cost
Google restricts Advanced to a $19.99 Google One AI Premium Plan, though you can get two free months of service before you have to pay.
You can't pay for Gemini Advanced by itself. Google One Premium subscribers ($9.99/month) must pay double for Gemini Advanced. You get all the standard Premium perks like 2TB of storage, a Google VPN, and Photos AI tricks like Magic Eraser.
Keep in mind that if you subscribe to a Google One AI Premium Family plan, only the main account user gets access to the AI; every other family member only receives benefits like extra storage and advanced Photos editing tools.
Ultimately, your choice on whether to pay for Gemini Advanced will depend on whether you use Gemini casually, professionally, or academically. You can get standard Gemini AI answers using the app without paying; Advanced gives you more rigorous tools.
Google Gemini (and Bard) controversy
Google Bard got off to a rocky, controversial start. While the Gemini AI rebranding is meant to shake off the bad moments, it, too, isn't without controversy.
During the initial Google Bard reveal, it gave the wrong answer to a question during its reveal, prompting some employees to state that Google Bard had been rushed out.
Another point of controversy came when a former Google engineer, Jacob Devlin, claimed that Google trained Bard on ShareGPT, a forum where people post the AI's responses. And an internal source claimed that Google only stopped doing this when Devlin criticized the practice.
Google itself says that "Bard is not trained on any data from ShareGPT or ChatGPT," so we can't be certain if this is true or not.
After Bard's launch, Google's "Godfather of AI" quit, noting the overall dangers of AI to spread misinformation and take away people's jobs. Google has laid off tens of thousands of workers in the past year, and cited AI policies when justifying its most recent layoffs in the engineering and ad sales teams.
Later, we learned that the people tasked with training and improving Bard are dealing with stressful working conditions, given just three minutes to determine how accurate a response is when that answer might require significant research.
The latest controversy is specific to Gemini, as Google swaps it in to replace Assistant. Gemini for Android users criticized missing features like podcasts, routines, and reminders. Gemini project lead Jack Krawczyk also pointed out issues like "preachy guardrails" and the need for an interpreter to understand coding responses. Of course, these kinds of issues can be expected for a major software launch; the question is how quickly Google will resolve them.
Michael is Android Central's resident expert on fitness tech and wearables, with an enthusiast's love of VR tech on the side. After years freelancing for Techradar, Wareable, Windows Central, Digital Trends, and other sites on a variety of tech topics, AC has given him the chance to really dive into the topics he's passionate about. He's also a semi-reformed Apple-to-Android user who loves D&D, Star Wars, and Lord of the Rings.
For wearables, Michael has tested dozens of smartwatches from Garmin, Fitbit, Samsung, Apple, COROS, Polar, Amazfit, and other brands, and will always focus on recommending the best product over the best brand. He's also completed marathons like NYC, SF, Marine Corps, Big Sur, and California International — though he's still trying to break that 4-hour barrier.