Skip to main content
Tinbox Hero Light

​
Welcome to Tinbox

Tinbox is a robust command-line tool designed to tackle the challenges of translating large documents using Large Language Models (LLMs). Unlike other tools, Tinbox excels in handling extensive document sizes and navigates around model limitations related to size and copyright issues, ensuring seamless and efficient translations.

​
Why Tinbox?

Tinbox solves critical problems that make document translation challenging:

No OCR Needed

Directly translates PDFs using vision-capable multimodal models (GPT-4o, Claude Sonnet, Gemini Pro) without requiring OCR that leads to formatting loss and errors.

Large Document Handling

Efficiently processes large PDFs and documents that would typically fail or timeout with traditional translation tools.

Overcomes Model Limitations

Bypasses common model refusals due to copyright concerns, document size limitations, and rate limiting through intelligent algorithms.

Quality & Consistency

Smart algorithms ensure consistent translations across document sections while maintaining context and terminology through glossary support.

​
Key Features

Smart Document Handling

Supports PDFs, Word documents (DOCX), and text files with intelligent processing for each format.

Intelligent Algorithms

  • Page-by-Page: Best for PDF documents
  • Context-Aware: Default for text with smart splitting
  • Sliding Window: For sequential long documents

Flexible Model Support

Works with OpenAI (GPT-4o, GPT-5), Anthropic (Claude), Google (Gemini), or local Ollama models.

Glossary Support

Maintains consistent terminology across documents with automatic term detection and custom glossaries.

Checkpoint & Resume

Handle large documents with automatic progress saving and the ability to resume interrupted translations.

Cost Tracking

Estimate translation costs and time before starting, with full progress and cost tracking during translation.

​
How It Works

1

Load Your Document

Tinbox automatically detects your document type (PDF, DOCX, or TXT) and uses the appropriate processor.
2

Smart Processing

Documents are intelligently split using context-aware algorithms that maintain meaning across chunks.
3

LLM Translation

Each chunk is translated using your chosen LLM provider while maintaining consistency through glossaries and context.
4

Quality Output

Receive your translated document with preserved formatting and consistent terminology throughout.

​
Use Cases

Translate technical manuals, API documentation, and software guides while maintaining consistent terminology with glossary support.
Process research papers and academic documents with context-aware translation that preserves meaning across sections.
Convert business presentations, reports, and proposals while maintaining professional formatting and terminology.

​
Quick Example

# Translate a text document to German
tinbox translate --to de --model openai:gpt-5-2025-08-07 document.txt

# Translate a PDF to Spanish (requires vision-capable model)
tinbox translate --to es --algorithm page --model openai:gpt-4o document.pdf

# Use local models with Ollama
tinbox translate --to fr --model ollama:llama3.1:8b document.txt
Ready to get started? Head over to the Installation guide to set up Tinbox, or jump straight to the Quick Start to see it in action.

​
Community & Support

Tinbox is an open-source project. If you encounter issues or have questions:
  • Report issues: Open an issue on our GitHub repository
  • Get help: Check the troubleshooting section in our docs
  • Contribute: We welcome contributions and improvements
Tinbox requires Python 3.12 or higher and works best with vision-capable models for PDF translation.

Build docs developers (and LLMs) love

Get started for freeTalk to us