Creating Gems with Google Docs

Author(s): Ime Inyang Jr.

Originally published on Towards AI.

In August 2024, Google released an interesting feature on Gemini – Google’s state-of-the-art Multimodal Large Language Model (MLLM) – to allow users to customize models on the platform. It is called a gem (As you might guess, short for “Gemini”.,

Gems was released to free subscribers in March 2025, making this tutorial useful for all parties involved.

In this tutorial, we’ll learn how to create simple gems for your different needs and how you can make them more powerful with Google Docs.

N/B: This tutorial assumes that you already have a functional Google Account and that you already use Gemini.

making a gem

gems

Gem is basically a Gemini model with a custom directive to guide its interaction with the user. These instructions are defined before interaction with the model begins. we call them system prompts,

System cues allow the language model to assume personality relevant to the user’s needs. This reduces the need to create chat history with a model to create some level of work efficiency.

These instructions are announced once and remain consistent throughout the conversation.

your first gem

step 1To create your first gem, go to: GeminiOpen the left panel with three parallel strokes and select “Explore Gems”.

This will navigate to your Gem Manager, where you’ll see the premade gems by Google and any you’ve created, if you’ve already created one.

step 2: Click on the “New Gem” button. This will transfer you to the laboratory area, where you can customize gems.

step 3: Define your gemstone name, description and instructions.

The directions are what the gems move on. This is where Prompt Engineering shines. Instructions can range from small tasks to a list of multimodal tasks. Great instructions are specific and broken down.

step 4: Test your Gem in the preview window on the right: Interact with your Gem and quickly see how it will react when published.

Step 5: Save your gem by clicking the “Save” button above the preview window.

You can now return to the Gemini homepage and find your gemstone listed.

Document Driven Gems

Knowledge

To increase the power of your gems, there is a feature in the Lab called “Knowledge”. Knowledge is where you can provide context to your gems to improve their potential. These can be images, documents, code or slides.

You can also describe workflows in documentation for your Gem to follow, as long as the processes fall within Gemini’s passive capabilities.

Let’s say you want to prepare a certain type of photos, instead of always uploading samples in every conversation, you can supply them in the knowledge area and instruct Gem to follow the examples in the knowledge.

career planner gems

Docs is my favorite because of its editing flexibility and its multimedia freedom. They can serve as a type of skill for your gems.

I created a career planner using Docs. This way:

1. Create a new document in Drive.
2. Document name and describe function: This is defined at the top of the document. This increases Jem’s confidence that he is using the correct knowledge. It also serves as a title for you, the builder, so you know what document you’re working with when you look at it again in the future.
3. Define the purpose of the document:This serves as a kind of extension of Jem’s instructions. Here, I list what is expected from GEM. If you expect the gems to follow a specific order, be sure to define it here.
4. Describe output data: A career planner is essentially a workflow: a series of steps and processes to arrive at a clear path. Each step has its own expected input/output, which feeds into either the next or previous step/process in the workflow. Here, I define the expected data and output formats for each step in the workflow.
5. Define Workflow: The workflow in this document is called “Questions” – a list of grouped questions that Gem asks the user to understand the user’s desired career direction.
6. write in markdown: To make sure Jem understands the flow and different levels in each step, I wrote the entire process in Markdown format.
7. Add document as knowledge: Return to your Gem’s lab area and select the “Add Files” button (,) under Knowledge and select “Add to Drive”. Select your recently created document from the pop up window. ,You can define and add as many documents as you need for your use case,
8. Examination: Interact with your gemstone in the preview window to make sure it meets your expectations. You can edit the instructions or your documentation until your Gem performs optimally.
9. save: When you are satisfied with the performance of your Gem, you can save it and start mapping your career journey.

Markdown is a lightweight markup language with a simple syntax that uses plain text formatting to create structured documents. It essentially uses hashtags (#) to specify titles. The more hashtags (##) the lower the heading level,

This motivational strategy was adapted from Anthropic Skill Functionality in Cloud AI (see). Here And Here,

There’s nothing you can’t do with a little quick engineering!

What makes it especially useful is that it’s all native to your Google account – no drag-and-drop required. And your gems are saved in your Google Drive.

You can also share your gems for others to use. Do this:

• Returning to the Gem Manager, click the Share button on Gem.
• Changing general access from “restricted” to “anyone with the link”.
• Setting the permission to “Viewers”.
• Copy the link and click “Done”. This will prompt you to grant viewer permission for the document/knowledge source. Click “Share”.

chat with Career Planning Gems

To get the full benefit of any knowledge, call it up in the instructions window and specify how and when the knowledge should be used. This ensures that your gem always refers to that knowledge when performing specified tasks. Failing to do so can lead to Jem becoming confused, especially when imparting much knowledge.

You can experiment with how Gem calls up knowledge, how much knowledge can be loaded per time, and the latency of Gem due to referencing its knowledge to ensure instructional consistency. This is important for user experience.

Given that Gemini is multimodal, there is almost nothing you can’t make or create in Gemstone.

You can also try creating a slide builder/generator, a photo review, or a movie maker. It all depends on your imagination, your well-designed workflow and your prompting skills.

Step right in and enjoy creating!

N/B:

The functionality of your Gem may be affected by your Google subscription plan, as higher plans provide advanced Gemini capabilities (see description). Here,

Published via Towards AI