Mercury · Python notebooks to web apps

Beautiful web apps. Built with Python.

You don’t need to be a frontend wizard. Add widgets to your notebook, and Mercury makes them work and look good.

Ready-made widgets and polished layouts
Live preview in MLJAR Studio and JupyterLab
Publish to the cloud or host it yourself

Open source · Apache-2.0 · View on GitHub

iPhone Price Explorer built with Mercury: filters, price summaries, and a Python chart

Add widgets. Mercury handles the interface.

Choose inputs, dropdowns, sliders, tables, charts, or chat widgets. Use them in Python and get a consistent interface without writing HTML, CSS, JavaScript, or callbacks.

Three notebook cells

# Cell 1
import mercury as mr

# Cell 2
name = mr.TextInput(label="What is your name?")

# Cell 3
mr.Markdown(f"## Hello {name.value}! 👋")

Try the interaction

What is your name?

Hello Ada! 👋

Change the name to see the output update.

Browser illustration. In Mercury, widget changes rerun the notebook cells below them.

Browse the widgets or follow the notebook-to-web-app guide.

See your app as you build it.

Keep your notebook and live app preview side by side. Edit your Python, try the widgets, and check the results without leaving your development environment.

Available in MLJAR Studio and JupyterLab with the Mercury extension.

1. Open your notebook.
2. Add Mercury widgets.
3. Open the live app preview and try it.

Watch the notebook and app together. The demo has no audio.

Your app, your style.

Start with a polished default, then make it yours. Choose a ready-made theme or set your colors, fonts, and branding in config.toml. Select a style below to see the same app with a different look.

Mercury prediction app with Corporate Blue styling, sidebar controls, prediction cards, and a chart — Corporate Blue · Same notebook, different lookGet this theme →

Explore all nine ready-made themes →

Ready to share? Publish your app.

Give people a link they can open in their browser. They don’t need Python, Jupyter, or your development environment.

Publish from Studio in one click

Use MLJAR Platform for managed hosting. After setting up your account and selecting your app files, click Publish in Studio. The platform builds the environment and deploys your app.

Include your notebook, dependencies, and data files. Keep working on your app while MLJAR handles the servers.

Deploy to MLJAR Cloud Cloud guide

Host it on your own infrastructure

Package your notebooks and dependencies with Docker. Run Mercury on your own server or inside your organization’s network, with control over the environment and data.

Mercury is open source and free to self-host. You provide and manage the infrastructure.

Self-host with Docker

Keep private apps private.

Set a password and share access with your team for internal dashboards, reports, and tools. Visitors enter your shared password before opening the app.

Set up password protection

Mercury login screen with a password field and Log in button protecting access to a private app

Built with Mercury.

Real data, real notebooks, and interfaces people can explore.

iPhone Price Explorer

Compare launch prices, adjust for inflation, and explore how screen sizes changed.

Open app Read the story View notebook

GitHub outages

Explore years of incidents with an activity calendar and filters for impact and component.

Read the story View notebook

Find your next data job

Filter hiring posts by skills and work preferences, and see the funnel update.

Read the story View notebook

Programming language trends

Compare language mentions across years of job posts with an interactive Sankey.

Read the story View notebook

Explore Python apps and guides →More Mercury examples →

Common questions

Do I need to know HTML, CSS, or JavaScript?

No. Add Mercury widgets in Python and use their values in your notebook. Mercury supplies the layout, styling, and interactions. You can customize the appearance when you want to.

Can I preview my app while I develop it?

Yes. Open the live app preview in MLJAR Studio or in JupyterLab with the Mercury extension. You can edit your notebook and test the app side by side. The live preview extension is not available in VS Code or Google Colab.

Can I use an existing notebook?

Yes. Add input widgets to your notebook and use their values in the cells below. Keep your Python analysis, charts, and outputs together, then serve the notebook as an app.

Can I change the colors and styling?

Yes. Use a ready-made theme or customize config.toml to change colors, fonts, branding, and other appearance settings.

Where can I deploy my apps?

Publish through MLJAR Platform for managed hosting, or self-host with Docker on your own server. The cloud service handles infrastructure; self-hosting gives you control of the environment.

Can I make my apps private?

Yes. Mercury supports shared-password protection. Visitors must enter the password before opening the app. User-based authentication is a separate paid option; contact MLJAR for details.

Is Mercury open source?

Yes. Mercury is available under the Apache-2.0 license and is free to self-host. You provide the hosting infrastructure, or choose managed hosting through MLJAR Platform.

Your next app starts with a notebook.

Install Mercury, add your first widget, and see what your Python can become.

Start building with Mercury View on GitHub

# Install Mercury
pip install mercury

# Start Mercury from the folder containing your notebooks
mercury

Run from the folder containing your notebooks to serve them as web apps.