Python notebooks on Android

A notebook built for a phone screen

Try these examples

Each screenshot is real output from the code beside it. Copy each block into its own cell and press Run.

import numpy as np
x = np.linspace(0, 2*np.pi, 200)
plot(x, np.sin(x))

Symbolic math with SymPy

Differentiate, integrate and solve, with results rendered as LaTeX. The ∑ Math Mode button inserts common SymPy functions, so there is less to type on a phone keyboard.

PYTHON · CELL 1

from sympy import symbols, diff, integrate, sin, exp
x = symbols('x')
diff(sin(x)*exp(x), x)

PYTHON · CELL 2

integrate(x**2 * exp(-x), x)

Mix in other languages

Each cell has its own language. Choose PL in the language selector to write Prolog. R, Bash, JavaScript, Lua and others work the same way, and every language can read and write the same files under /shared.

PROLOG · CELL 1

parent(tom, bob).
parent(bob, ann).
grandparent(X, Z) :- parent(X, Y), parent(Y, Z).

PROLOG · CELL 2

grandparent(tom, Who).

Run these as two cells. A cell that defines rules is loaded as a program, so the query goes in its own cell.

Bring your notebooks with you

SciREPL reads and writes the standard Jupyter format, so work moves between your phone and a laptop.

• Open .ipynb files with ☰ Menu → Import File. Saved outputs (text, images, LaTeX, tables) show straight away, without re-running anything.
• Export back to .ipynb with outputs included, so the notebook still displays in Jupyter or on GitHub.
• Export a readable document as HTML, Markdown, PDF, DOCX or LaTeX, with highlighted code, plots and equations.
• Also imports .py and .pl source files, CSV data, and SciREPL's own .srwb workbooks.
• Keep several notebooks open in tabs. They save automatically and come back when you reopen the app.
• Send a notebook anywhere with Android's share sheet.

Free or Pro?

Both editions run the same notebook. Pro bundles every runtime for offline use and adds the AI features.

What it is not built for

Worth knowing before you install:

• Python runs as WebAssembly, so heavy numerical work is slower than native Python on a laptop.
• %pip install works for pure-Python packages from PyPI. Packages with compiled C extensions need a prebuilt WebAssembly version, and many do not have one.
• There is no GPU, so it is not for training machine-learning models or processing very large datasets.
• Installing extra Python or R packages needs a network connection. The bundled runtimes do not.
• It is single-user: there is no real-time collaboration or shared server.
• Requires Android 7.0 or later. There is no iOS version; Pro also has an unsigned Windows preview.

Get started

1. Install SciREPL Pro, or the free edition, from Google Play.
2. Open the app. The input bar is at the bottom: pick a language on the left (Py, R, PL…), type some code, and press ▶ Run.
3. Paste one of the examples above to check that everything works.
4. To open an existing notebook, use ☰ Menu → Import File and choose a .ipynb file.
5. Tap ? in the header for the built-in help. Every help example can be copied, or inserted as a new cell.