You Can Now Create 'Gems' in Google's Gemini Chatbot

It’s been almost a year since OpenAI released its customizable GPTs, which are personalized versions of ChatGPT tailored for specific tasks. For example, ‘Dall-E’ is a custom GPT designed to generate AI images, while a “Creative Writing Coach” GPT offers feedback and guidance to help improve your writing skills.

Today, and Google has rolled out a similar feature within its Gemini ecosystem, and they are calling it ‘Gems.’

Gems were first previewed at Google’s I/O 2024 developers’ conference in May and are now available to Gemini Advanced, Business, and Enterprise users in over 150 countries. These Gems are accessible on both desktop and mobile devices, aiming to give users more control over their AI interactions.

What are Gems?

Gems allow you to customize Gemini to create your own personal AI experts on any topic you can think of. For instance, if you want a chatbot optimized to generate ideas for a product, you simply create a Gem and instruct it to brainstorm five ideas whenever prompted. You can create any Gem you want, whether it be a writing assistant, image generator, gym buddy, recipe creator, and more.

Essentially, you create a “Gem” by defining its behavior and functions, then deploy it to help with whatever task you have in mind.

Gems will be a part of the Gemini Advanced feature set, utilizing the Gemini Advanced model, which is currently at version 1.5 Pro. This model promises enhanced capabilities, making the customization experience richer and more intuitive.

How to Create a Gem

When you open Google Gemini, you’ll see a list of your Gems on the sidebar.

To create a Gem, click on the “Gems Manager” menu on the sidebar.

Inside the Gems Manager, you can choose from Google’s premade Gems or create your own from scratch by clicking on the “New Gem” button.

Give it a name and and provide instructions on how it should function. In the example below, I created a Gem that will create a 2–3 sentence summary of an article and generate SEO tags.

Instructions: Analyze the article and generate a 2–3 summary description. Also, generate 5 tags related to the article’s content.

Once you hit save, Gemini takes those instructions and configures the Gem to your specifications. If everything goes smoothly, you’ll see a success message.

From there, you can start using the Gem by clicking on the “Start Chat” button. It’s that simple. Check the example below where I uploaded a doc file about the top 5 AI innovations in 2024.

Awesome! It followed exactly my instructions and the results are really good. If you’re not satisfied with the Gem’s performance, you can always tweak the instructions to get better results.

Google’s Premade Gems

If you prefer not to create a Gem from scratch, you can explore Google’s set of premade Gems designed for various scenarios:

• Brainstormer: Provides inspiration, from fresh ideas for a themed party to the perfect gift for an upcoming birthday.
• Career Guide: Unlocks your career potential with detailed plans to refine your skills and achieve your career goals.
• Coding Partner: Levels up your coding skills and assists you in building projects while you learn.
• Learning Coach: Helps break down complex topics, making them easier to understand.
• Writing Editor: Evaluates your writing through clear, constructive feedback on everything from grammar to structure.

Can You Share Gems?

No, unfortunately, you cannot share Gems directly with others in Gemini at this time. Gems are currently designed for personal customization and use, which is a bit of a letdown.

Imagine the possibilities if users could share Gems with friends, colleagues, or even the wider community. This could open up a whole new world of collaboration, where you could tap into Gems created by others to enhance your own workflow.

How Much Does it Cost?

In terms of pricing, Gemini Advanced is priced at $20 per month. This gives you the following benefits:

• Access to the next-generation model, Gemini 1.5 Pro
• Priority access to new features
• Experience a 1 million token context window. This allows you to process and understand significantly more information — up to 1,500 page PDFs — so you can tackle more complex problems than ever before

The subscription also provides 2 TB of Google storage, which is a compelling offer, especially for users who need significant cloud storage.

Honestly, I still prefer using ChatGPT because of it’s more capable than Gemini. The storage is the only thing that prevents me from canceling my Gemini Advanced subscription.

Final Thoughts

It’s refreshing to see Google launching Gems, albeit several months later than OpenAI. While it took some time for Google to catch up, the wait seems worth it based on initial tests. Gems simplify tasks, automate repetitive prompts, and generally make my life a bit easier — though still not better than GPT-4o.

However, it’s important to keep in mind that OpenAI has a nearly year-long head start. Their GPT store now hosts millions of custom GPTs, which gives them a significant advantage in terms of variety and community engagement. It’s going to take some time for Google to build a similar ecosystem around Gems, and that might be one of the biggest challenges they face.

Additionally, with the announcement of Gems, Google is also rolling out Imagen 3, its latest image generation model, to Gemini apps. This new model promises to enhance the quality of images generated by Gemini, and I am eager to see how it stacks up against models like Midjourney V6.1 and Flux Pro.

‍