Skip to content

panel-live

Write, edit, and run Python interactively in the browser — no server required.

Turn any web page into an interactive Python playground with the <panel-live> web component. Works with matplotlib, pandas, scikit-learn, Panel, and 200+ packages from the Python ecosystem. Visualizations, analyses, dashboards, and interactive tools are fully interactive — users can view, explore, edit code, and re-run, all directly in the browser via Pyodide — no backend, no deployment, no infrastructure.

Try it

App Mode

The default mode renders your Panel app directly — no editor, just the output.

import panel as pn pn.extension(sizing_mode="stretch_width") slider = pn.widgets.FloatSlider(name="Value", start=0, end=10, step=0.1, value=5.0) pn.Column( "# Slider Demo", slider, pn.bind(lambda v: pn.pane.Markdown(f"**Current value:** {v:.1f}"), slider), ).servable()

Works with any Python

Not just Panel — run any Python code directly in the browser. Here's a matplotlib visualization with no Panel imports at all:

import matplotlib matplotlib.use("agg") import matplotlib.pyplot as plt import numpy as np x = np.linspace(0, 2 * np.pi, 200) fig, ax = plt.subplots(figsize=(6, 3)) for n in range(1, 5): ax.plot(x, np.sin(n * x) / n, label=f"sin({n}x)/{n}") ax.set_title("Harmonic Series") ax.legend(loc="upper right") ax.grid(True, alpha=0.3) fig

Click the Code button to edit the code and rerun it live in your browser!

Try changing the code to:

ax.set_title("Very Harmonic Series")
ax.legend(loc="upper left")

Perfect for library developers and educators who want to make it easy to explore, learn, and have fun.

Editor Mode

Edit the code below and press Run to see your changes.

import panel as pn pn.extension(sizing_mode="stretch_width") name = pn.widgets.TextInput(name="Your Name", value="World") count = pn.widgets.IntSlider(name="Repeat", start=1, end=5, value=1) def greet(name, n): msg = f"Hello, {name}! " * n return pn.pane.Markdown(f"## {msg.strip()}") pn.Column(name, count, pn.bind(greet, name, count)).servable()

Playground Mode

A side-by-side editor and live preview. Edit the code on the left and press Run.

import panel as pn pn.extension(sizing_mode="stretch_width") a = pn.widgets.FloatInput(name="A", value=10, step=1) b = pn.widgets.FloatInput(name="B", value=3, step=1) def compute(a, b): return pn.pane.Markdown(f"## {a} + {b} = **{a + b:.4g}**") pn.Column( pn.Card( pn.Row(a, b), pn.bind(compute, a, b), title="Mini Calculator", ), ).servable()

Playground

Explore the full-screen Playground — a curated collection of interactive examples you can edit and run instantly.

API Editor

Explore and configure every <panel-live> attribute interactively with the API Explorer.

Features

  • 3 modes: app (output only), editor (code + output), playground (side-by-side)
  • Web Worker execution — Pyodide runs in a Dedicated Worker, keeping the page responsive
  • Light / dark / auto theming that follows the host page
  • CSS custom properties for full branding control
  • Real-time print output — print() streams incrementally as code executes
  • Multi-file support via <panel-file> child elements
  • Explicit requirements via <panel-requirements>
  • MkDocs integration via fenced code blocks and pymdownx.superfences
  • No server needed — runs entirely in the browser via Pyodide

Known Limitations

Because panel-live runs entirely in the browser via Pyodide (WASM), some features that require a live server are not available:

  • No .plot() calls — matplotlib's plt.show(), plotly's fig.show(), and similar display methods do not work. Instead, return the figure object as the last expression (e.g. fig) and panel-live will render it.
  • No .show() calls — Bokeh's show(), Panel's .show(), and similar server-dependent methods are not supported.

For a full list, see Known Limitations.

Get Started

pip install panel-live