T+ Thonny-PlusDownload
Menu

OPEN-SOURCE PYTHON IDELOCAL-FIRST AI

Python learning,
with a little more intelligence.

A familiar Thonny experience with a local, open-source AI assistant that explains code, suggests fixes, and helps you learn without taking over.

Download for Windows See how it works
v0.1 community build · local AI
Thonny-Plus running a Python program with a NameError while the local AI assistant explains the variable-name mistake and offers a hint01 · Python editor02 · Local AI guidance
REAL PRODUCT VIEWA beginner sees the error, the explanation, and the next useful hint in one workspace.
01
LOCAL MODEL

Your code stays on your machine.

02
BEGINNER-FIRST

Guidance that teaches, not shortcuts.

03
OPEN SOURCE

Built in the open, ready to extend.

WHY THONNY-PLUS

A guide beside you.
Not a pilot above you.

AI assistance should make the next idea clearer—not make the learner disappear from the loop.

01✦

Ask about your code

Run a program and get a plain-language explanation grounded in the code and error you can already see.

Why did I get a NameError?→
02↺

Get one useful hint

Ask for a nudge instead of a finished answer. The learner stays in charge of the next edit.

numbernumberr
03⌂

Keep learning local

The assistant runs with a supported open model on the learner’s computer—private by design.

LOCAL AI READY PRIVATE

LEARNER WORKFLOW

Run. Understand.
Try again.

Thonny-Plus keeps assistance close to the editor and shell. No tab switching, no copied error messages, and no loss of context.

  1. 1
    RUN

    Execute the program

    Thonny displays the traceback and highlights the relevant line using its familiar beginner-friendly tools.

  2. 2
    UNDERSTAND

    Read the explanation

    The assistant connects the traceback to the learner’s code and explains the underlying Python concept.

  3. 3
    RETRY

    Ask for one hint

    The learner gets a focused next step, edits the code, and runs it again—keeping the feedback loop intact.

IMPLEMENTATION

Thonny at the core.
AI at the edge.

The assistant is added as a focused side panel. It reads the active editor context and latest runtime error, sends only the necessary prompt to a local model, and returns learning-first guidance.

01EDITOR + SHELL

Capture context

The active Python source, selected line, and latest traceback form the working context.

source + traceback
02CONTEXT LAYER

Shape the prompt

A learner-first instruction asks for explanation, concept, and one useful hint—not a full solution.

explain → teach → hint
03LOCAL MODEL

Generate privately

An open-source model runs on the machine and produces a response without uploading the project.

localhost / model
04ASSISTANT PANEL

Return guidance

The result appears beside the code with privacy status and an action for the next hint.

explanation + next step
PRIVACY BOUNDARYThe local model is the boundary: source code and tracebacks stay on the learner’s device.

SYSTEM REQUIREMENTS

Ready for a modern
learning machine.

The local AI model needs more resources than standard Thonny. These baseline requirements keep the editor and assistant responsive together.

RECOMMENDED16 GB RAM gives the local model more room and smoother response times.
Processor
Intel Core i5 or equivalent 64-bit CPU
Memory
8 GB RAM minimum · 16 GB recommended
Storage
10 GB free space for the app, model, and updates
Operating system
64-bit Windows 10/11 or a modern 64-bit Linux distribution
Display
1280 × 720 minimum · 1920 × 1080 recommended
Graphics
No dedicated GPU required · compatible GPU acceleration optional
Connection
Internet for first setup and model download; offline use afterward

GET THONNY-PLUS

Download it.
Learn locally.

Install the current Windows 64-bit classroom build directly, or browse the source and releases on GitHub.

Download Windows installer View source on GitHub Thonny Classroom 6.0.0 · Classroom.1 · Windows x64

OPEN BY DEFAULT

Fork it.
Teach with it.
Make it yours.

Thonny-Plus builds on the spirit of Thonny: small, welcoming, and transparent. The assistant is designed around replaceable open models, clear prompts, and community-shaped learning.

Open the repository Meet upstream Thonny ↗
THONNY-PLUS / LOCAL FLOW● ● ●
1 capture editor + traceback
2 build learner-first prompt
3 run open model locally
4 render explanation + hint

✓ IDE ready
✓ Local AI connected
✓ Source remains private
PYTHONLOCAL AICOMMUNITY

READY WHEN YOU ARE

Less mystery.
More “I get it.”

Download for Windows