Google Labs
Google Labs
Try Google's new AI tools early; many things are still experimental.
Google's home for AI experiments, where you can try new tools early before they become official products.
Overview
Google Labs is where Google presents new AI experiments and brings successful ones into official products. The labs.google site gathers experimental projects such as Opal (create AI mini-apps in natural language), Jules (an asynchronous coding assistant), Stax (a toolkit for evaluating models and prompts) and Playground (create games from text descriptions).
Many familiar products began here and have "graduated", such as Gemini (formerly Bard), Google AI Studio (formerly MakerSuite) and Google Flow (formerly VideoFX). So following Google Labs is a way to learn early where Google is heading in AI.
For businesses, this is a place to find experimental tools that can help automate small tasks or create content. Note that the projects are experimental, so they may change, have limited availability or be discontinued.
Highlights
One place for Google's AI experiments
A single page listing Google's new AI projects with short descriptions.
Early access
Many experiments are open for users to try and give feedback before official launch.
No-code tools
Opal lets you create, edit and share AI mini-apps from a description alone.
Tools for technical people
Jules helps fix bugs and write tests; Stax helps evaluate models and prompts.
See the product roadmap
The graduated products section shows which experiments have become official products.
What it's for
Creating internal mini-apps
Office staff use Opal to build a small app, for example turning notes into a summary email, without coding.
Supporting the dev team
The engineering team tries Jules for repetitive tasks such as small bug fixes or adding tests.
Checking prompt quality
People building AI workflows use Stax to compare and evaluate prompts before putting them into operation.
Creating creative content
Marketing follows video and music creation tools such as Google Flow to try ad ideas.
Keeping up with trends
Business owners visit the site periodically to see what Google is experimenting with.
How to use it
- 1
Open the homepage
Go to labs.google to see the list of current experiments.
- 2
Browse the Experiments section
Read each project's short description and choose a tool close to your needs.
- 3
Sign in when asked
Many experiments need a Google account; use an account that fits your company policy.
- 4
Try with sample data
Test the tool with non-sensitive content and read each experiment's terms.
- 5
Follow updates
Use Stay connected to get news when new experiments appear.
Video guides
Celebrating the AI experiments community - labs.google
Introducing Opal
Google for Developers
How to get started on Opal
Google for Developers
Real-world example
A small travel company wants its consultants to draft itineraries faster. The team lead goes to Google Labs and tries Opal to build a mini-app: enter a destination, number of days and budget, and the app returns a draft itinerary. After testing internally with sample requests, the team decides to use it as a drafting aid, with staff still checking before sending to customers.
Tips & notes
Treat it as experimental
Do not make an important process depend entirely on a project that may change or be discontinued.
Check availability
Some experiments may not be open to every country or every type of account, so check when you visit.
Read the data terms
Each experiment has its own terms; avoid entering customer data until it is clear.
Look at the graduated section
If you need stability, prefer products already officially launched such as Gemini or Google AI Studio.
FAQ
Is Google Labs free?
The showcase site is free; each experiment has its own terms of use, see each project's official page.
Can the experiments be used in Vietnam?
It depends on the project; the homepage does not state country coverage, so you need to try accessing directly.
Will experiments last long term?
Not certain; some become official products, others may change or stop.
I do not know how to code, what should I try first?
Opal is an easy place to start because it creates mini-apps entirely from natural-language descriptions.
Useful links
Want to apply this in your business?
2dm looks at your workflows, picks the right tools and sets them up with your team. The first call is free.
Same topic
There's An AI For That
4.0
A continuously updated directory and news feed of AI tools, helping you find the right tool for the job you need done.
Hugging Face
4.5
The central hub of the open AI community: models, datasets and demo apps (Spaces) you can try right in the browser.
Product Hunt
3.5
Where new tech products launch every day and are ranked by the community, many of them AI tools.