README.md

metadata

title: LogosAI
emoji: 📚
colorFrom: purple
colorTo: indigo
python_version: 3.13
sdk: gradio
sdk_version: 5.33.0
app_file: app.py
pinned: false
tags:
  - mcp-server-track
license: mit
short_description: Deeply read any text in any language, from news to philosoph

LogosAI 🚀

An intelligent system that reads, reasons about and transforms language.

Deeply read any text in any language — OCR, correct, interpret & translate in one interactive demo!

---

Why LogosAI?

• All-in-one pipeline: Upload images/PDFs/plain text → OCR → AI-powered proofreading → genre-aware interpretation → translation.
• Novel hackathon demo: Leverages Mistral for OCR & correction, Gemini for interpretation/translation, served live as an MCP server.
• Genre intelligence: Tailored system prompts for news, narratives, poetry, philosophy and general texts.

---

⚙️ Features

1. OCR & Correction
   – Extract text from images/PDFs/text files
   – Proofread markdown output with AI for typos & formatting
2. Interpretation
   – Deep analysis of syntax, rhetoric & cultural context
   – Supports “General” & “News” modes (more coming soon!)
3. Translation
   – Faithful transforms into Deutsch, English, Français, Русский or 中文
4. Live Gradio + MCP
   – runs with mcp_server=True

---

🚀 Quick Start

1. Clone this repo
2. pip install -r requirements.txt (with python 3.13!)
3. Set your keys:

   export MISTRAL_API_KEY="…"
   export GEMINI_API_KEY="…"

4. Launch the app:

   python app.py

5. Open the Gradio link and start uploading, interpreting, translating!