Introduction to Gemini

Google is making a significant push into the world of generative AI with its flagship suite of models, apps, and services called Gemini. But what exactly is Gemini, and how can you use it? How does it compare to other generative AI tools like OpenAI’s ChatGPT, Meta’s Llama, and Microsoft’s Copilot? To help you stay up-to-date with the latest developments, we’ve put together this comprehensive guide, which will be regularly updated with new information on Gemini models, features, and news.

What is Gemini?

Gemini is Google’s next-generation generative AI model family, developed by its AI research labs, DeepMind and Google Research. It comes in four flavors:

1. Gemini Ultra: A very large model
2. Gemini Pro: A large model, with the latest version being Gemini 2.0 Pro Experimental, which is Google’s flagship
3. Gemini Flash: A speedier, "distilled" version of Pro, with a smaller and faster version called Gemini Flash-Lite, and a version with reasoning capabilities called Gemini Flash Thinking Experimental
4. Gemini Nano: Two small models, Nano-1 and Nano-2, which can run offline

All Gemini models are natively multimodal, meaning they can work with and analyze more than just text. They were pre-trained and fine-tuned on a variety of public, proprietary, and licensed audio, images, and videos, as well as codebases and text in different languages.

Gemini Apps and Models

Gemini is separate from the Gemini apps on the web and mobile (formerly Bard). The Gemini apps are clients that connect to various Gemini models and layer a chatbot-like interface on top. You can access Gemini on the web, Android, and iOS devices.

Gemini Advanced

Gemini Advanced is a premium version that offers additional features, such as priority access to new features, the ability to run and edit Python code directly in Gemini, and a larger "context window." Gemini Advanced users can also access Google’s Deep Research feature, which generates research briefs, and a memory feature that uses old conversations as context.

Gemini Across Google Services

Gemini is available across various Google services, including Gmail, Google Docs, Slides, Sheets, Drive, and Meet. It can also be used in Google’s database products, cloud security tools, and app development platforms.

Gemini Extensions and Gems

Gemini Advanced users can create custom chatbots called Gems, which can be generated from natural language descriptions and shared with others. Gemini extensions allow the Gemini apps to tap into Google services, such as Google Drive, Gmail, and YouTube.

Gemini Live

Gemini Live allows users to have in-depth voice chats with Gemini, which can be accessed on mobile devices and the Pixel Buds Pro 2. It enables users to interrupt Gemini while it’s speaking and adapt to their speech patterns in real-time.

Image Generation via Imagen 3

Gemini users can generate artwork and images using Google’s built-in Imagen 3 model, which can more accurately understand text prompts and produce fewer artifacts and visual errors.

Gemini for Teens

Google introduced a teen-focused Gemini experience, which allows students to sign up via their Google Workspace for Education school accounts. This version has additional policies and safeguards, including a tailored onboarding process and an AI literacy guide.

Gemini in Smart Home Devices

Gemini is being integrated into various Google-made devices, such as the Google TV Streamer, Pixel 9 and 9 Pro, and the newest Nest Learning Thermostat. It will soon be able to summarize security camera footage from Nest devices.

What Can the Gemini Models Do?

Gemini models can perform a range of multimodal tasks, such as transcribing speech, captioning images and videos, and generating text. The different tiers of Gemini have varying capabilities, including:

• Gemini Ultra: Can be used for tasks like physics homework, solving problems step-by-step, and identifying scientific papers relevant to a problem
• Gemini Pro: Excels in coding performance and complex prompts, and can take in up to 1.4 million words, two hours of video, or 22 hours of audio
• Gemini Flash: Can natively generate images and audio, and is faster than Gemini’s previous generation of models
• Gemini Nano: Can run on devices like the Pixel 8 Pro, Pixel 8, Pixel 9 Pro, and Pixel 9, and powers features like Summarize in Recorder and Smart Reply in Gboard

Pricing

The Gemini models are available through Google’s Gemini API, with free options and pay-as-you-go pricing. The base pricing for each model is as follows:

• Gemini 1.5 Pro: $1.25 per 1 million input tokens (for prompts up to 128K tokens) or $2.50 per 1 million input tokens (for prompts longer than 128K tokens)
• Gemini 1.5 Flash: 7.5 cents per 1 million input tokens (for prompts up to 128K tokens), 15 cents per 1 million input tokens (for prompts longer than 128K tokens)
• Gemini 2.0 Flash: 10 cents per 1 million input tokens, 40 cents per 1 million output tokens
• Gemini 2.0 Flash-Lite: 7.5 cents per 1 million input tokens, 30 cents per 1 million output tokens

Project Astra

Project Astra is Google DeepMind’s effort to create AI-powered apps and agents for real-time, multimodal understanding. While it’s still in the project stage, demos have shown its potential for simultaneous processing of live video and audio.

Is Gemini Coming to the iPhone?

There’s a possibility that Gemini might come to the iPhone, as Apple has said it’s in talks to put Gemini and other third-party models to use for features in its Apple Intelligence suite. However, no official announcement has been made yet.