Python 3.11+ · source application
Cicero
Translate documents with separate translation, review, correction, and audit steps. Saved progress lets a provider error interrupt the work without erasing it.
What it does
- Reads text, DOCX, PDFs, and supported images; scanned material can use local Tesseract OCR.
- Saves a resumable workspace and produces UTF-8 plain text, with optional DOCX and PDF exports.
- Separates fluency review from comparison against the source.
Compatibility
Requires Python 3.11+ and a supported provider API key. Provider charges apply.
The source package includes Windows Install_Cicero.cmd and Start_Cicero.cmd launchers. A separate executable is not published.
OCR requires a separately installed Tesseract program; its Python wrapper does not include the OCR engine.
Download & installation
Available as a documented source installation. The download contains the public repository's source files.
Source code · requires the setup described below · not a prebuilt executable
- Download and extract the source ZIP, keeping the files together in a writable folder.
- On Windows, run Install_Cicero.cmd, then Start_Cicero.cmd. Alternatively install requirements.txt and run python interactive_translator.py.
- Choose a document, target language, and models. Enter your provider key at the program's hidden prompt.
Limitations
Translation does not preserve the source layout. OCR, table structure, paragraph grouping, and translation accuracy need review; interrupted provider work may require another paid call.