hydropix/TranslateBooksWithLLMs

What it solves

Translating long-form content like books, subtitles, and documents is often difficult because of file size limits, loss of formatting, and inconsistent terminology across thousands of pages. This tool automates the translation of large files while maintaining the original structure and stylistic consistency.

How it works

The application uses an intelligent chunking system to break large documents into manageable segments for LLMs, preserving context between them. It supports both local models (via Ollama or OpenAI-compatible servers) and various cloud providers (OpenAI, Gemini, DeepSeek, etc.). To ensure quality, it can automatically derive glossaries and writing styles from the text or allow users to apply custom style presets extracted from other books.

Who it’s for

It is designed for users who need to translate entire books or long documents (EPUB, SRT, DOCX, TXT) and want to maintain professional formatting and a consistent narrative voice without manually splitting files.

Highlights

• Format Preservation: Keeps EPUB styles, structure, and SRT timecodes exactly as they were in the original.
• Unlimited Length: Handles documents of any size through a checkpoint system that allows users to resume interrupted jobs.
• Flexible Provider Support: Works with local LLMs for privacy or cloud APIs for speed and scale.
• Style & Glossary Tools: Features automatic glossary generation and the ability to extract and reuse specific writing styles for consistent register and imagery.
• Parallel Processing: Supports concurrent chunk translation for cloud providers to speed up the process.