Google has no shortage of apps and services, but some of its most interesting ideas are hiding in plain sight. Tucked away in Google Labs are experimental features and projects that don't get nearly as much attention as the company's big launches.

I started digging through them expecting to find a few gimmicks, but came across several experiments that are surprisingly useful.

Some solve small everyday annoyances, while others offer a glimpse at where Google's apps and AI could be headed next.

Gemini Notebook made me rethink Google Labs

It was the first Labs experiment I kept using

I had always thought of Google Labs as the place where Google puts its interesting ideas before deciding what deserves a permanent spot in its products.

Some experiments are fun to play with, but I don't always find a reason to return to them. Gemini Notebook changed that for me.

Google originally launched it as NotebookLM in 2023 as part of an experimental project called Project Tailwind.

The idea was simple: instead of asking an AI to answer questions about whatever it could find, I could give it my own sources and have it work from that material.

Google renamed NotebookLM to Gemini Notebook in July 2026 and is connecting notebooks more closely with the Gemini ecosystem.

Notebooks can now sync between Gemini and Gemini Notebook, allowing me to organize chats and sources in Gemini while utilizing NotebookLM's specialized research tools.

For me, that's the interesting part about Google Labs. I went in expecting to find disposable experiments and instead found something that had grown into a useful research tool.

Gemini Notebook reminded me that some of Google's best ideas start in Labs, get better over time, and eventually become tools I want to keep using.

Opal showed me I could build without knowing how to code

It made coding feel less like a prerequisite

I've played around with code enough to understand the basics, but building a useful app from scratch is still a different story. That's why Opal caught my attention.

I can describe what I want an app to do, and Opal turns that description into a workflow. I can see how the different steps fit together, change them, and refine the result without writing the underlying code.

Google also added an agent step that can decide which tools and models to use based on my goal, ask follow-up questions when it needs more information, and even remember details across sessions.

That makes experimenting feel much more accessible. Instead of thinking, "I don't know enough code to build this," I can start with an idea and figure out how far I can take it.

And because Opal handles the hosting, I can use and share what I create rather than leaving it as a half-finished experiment.

Stitch made designing an app feel much less intimidating

It helped me turn an idea into a usable design

Designing an app can feel harder than it looks. I can have a clear idea of how I want something to work, but turning that idea into a polished interface is another challenge.

Stitch makes that first step easier by letting me describe an app in plain language and turning the idea into a high-fidelity UI.

I can start with an image or a rough sketch and create multiple versions, refining the design instead of committing to the first result.

Stitch also turns those screens into interactive prototypes, helping me better understand how the app will flow.

I don't need to know exactly how to design every screen before I begin. I can experiment, change my mind, and let Stitch handle the more tedious aspects.

Mixboard gives me a better way to explore ideas visually

Giving my ideas room to take shape

Some ideas are difficult to explain with words alone. I might know roughly what I want something to look like, but finding the right images and putting them together can take more effort than the idea itself.

Mixboard gives me a visual canvas where I can collect images, generate new ones, add text, and keep changing the direction as I go.

What makes it more interesting than a traditional mood board is how easily I can experiment. I can start with a prompt or my own images, ask Mixboard to generate different visuals, and use natural-language instructions to change what's already on the board.

If I like an idea but want to see a few variations, I can regenerate it or ask for something similar instead of starting over.

That makes Mixboard useful even when I don't have a clear result in mind. I can throw a few ideas onto the canvas and see what they turn into.

Google's learning experiments are more interesting than they sound

They aren't only for students

I expected Google Labs' learning experiments to be aimed mostly at students, but I can see myself using a few even when I'm not studying for an exam.

Its current Labs lineup includes learning-focused experiments alongside tools for creating and exploring.

Learn Your Way, for example, can turn textbook material into interactive lessons, quizzes, narrated slides, audio lessons, and mind maps. It also adapts the experience to the learner's grade level and interests.

Then there's Illuminate, which takes a completely different approach. It can turn research papers into short audio conversations between two AI-generated voices, making dense academic material easier to consume when I don't necessarily want to sit down and read a paper from beginning to end.

Google Labs made me curious about what Google might build next

What surprised me most about Google Labs was seeing how many different directions Google is exploring. Opal lets me turn an idea into a working mini-app, Stitch makes it easier to experiment with app designs, and Mixboard gives me a visual space to develop ideas.

Some Labs ideas may eventually become full products while others won't. Still, the website is worth visiting because you can try these ideas while they're still taking shape.