A user-friendly, Python-based editor that leverages a locally hosted LLM (via Ollama) for fast text enhancements, proofreading, and rewriting. Whether you're polishing a blog post, correcting grammar, or simplifying your wording, this editor lets you use local generative AI right from your desktop — no cloud services needed.
1. Inline Editing and Diff View
- Type or paste your plain text into the editor, then click an editing action (Grammar, Proofread, Rewrite, etc.).
- The application compares your text against the AI-edited output and highlights changes inline.
- Deletions appear in red, additions in green.
2. Multiple Editing Modes
- Built-in prompts let you quickly fix grammar, streamline awkward phrases, or make your text more concise.
- A customizable "Custom Prompt" option allows you to craft specialized instructions for the LLM.
3. Model Selection (v0.12)
- Choose any locally downloaded Ollama model from the Model dropdown at the top.
- Click Refresh to reload the list after adding/removing models in Ollama.
4. Accept / Reject All Changes
- Quickly apply all AI-suggested edits with "Accept All Changes," or revert them entirely with "Reject All Changes."
5. Scratchpad Logging
- Every edit is logged to a Markdown scratchpad file.
- The original user text and the LLM's edited text are saved side by side, with removed/added words highlighted in bold Markdown, helping you track changes over time.
6. Local LLM Support
- Connects to a locally hosted LLM through Ollama.
- No internet connection required — your content stays on your own machine.
1. Install Requirements
- Ensure Python 3.7+ is installed.
- Install the Ollama Python package:
pip install ollama. - Pull a model in Ollama
- If you have a GPU with 6 to 8 gb of VRAM :
ollama pull llama3.1:8b - If you don't have a GPU:
ollama pull qwen3:1.7b.
- If you have a GPU with 6 to 8 gb of VRAM :
2. Using uv (optional if installed)
cd TextEnhanceAI- Create venv:
uv venv - Install deps:
uv pip install -r requirements.txt - Run:
uv run python TextEnhanceAI.py
3. Run the App
- Launch with:
python TextEnhanceAI.py - If Ollama is running, you'll see the main editor window.
4. Use the Editing Buttons
- Type or paste your text in any language supported by your LLMs.
- Click an editing button (e.g., "Grammar", "Proofread").
- The edited text is displayed inline, with changes highlighted.
- Choose the LLM from the Model dropdown (defaults to
llama3.1:8b). Use the Refresh button to reload locally available models from Ollama.
Email: wenrolland@designecologique.ca
Version 0.12: Adds Model dropdown to choose local Ollama models and improves status text.
Version 0.11: Fixes formatting to keep line breaks after editing.