#4444th of 50 in Data and scraping for AIFor image, video and design, video and animation
video-subtitle-extractor
Desktop tool that OCRs hard-coded video subtitles into SRT files
- Stars
- 9.7k
- License
- Apache-2.0
- Last commit
- Apr 2026
- Last release
- Apr 2026
- Language
- Python
8,566 on 20 Apr 20269,722 on 9 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Extracts burned-in subtitles from video by sampling key frames, locating text regions, running local PaddlePaddle-based OCR, and removing duplicate lines to write an srt or txt file. It has a GUI (gui.py) and a CLI (backend/main.py), supports batch extraction, and covers 87 languages. Fast, auto and accurate modes trade speed against missed lines and typos; CUDA, DirectML and CPU backends are supported.
Who it is for: Users who need SRT files from hard-subbed videos, offline, on their own machine
Strengths
- OCR runs locally; no online OCR API keys or services needed
- Supports 87 languages including Chinese, Japanese, Korean, Arabic and Russian
- Batch extraction, plus typoMap.json to replace or strip specific recognized text
- Runs on CPU, NVIDIA CUDA, or DirectML; prebuilt release archives available
Weaknesses
- Accurate mode is described as very slow; fast mode may drop some subtitles
- Paths to videos and program must avoid Chinese characters and spaces
- ONNX path for macOS/ROCm is untested and the README says not to file issues
- README is Chinese-first and CUDA setup targets 11.8; RTX 50 series needs DirectML
What it needs
- GPU optional
- Needs PaddlePaddle 3.3.1, VideoSubFinder, CUDA 11.8 and cuDNN 8.6.0 (NVIDIA GPU), ONNX Runtime DirectML (AMD/Intel GPU)
- Models: PaddleOCR-based mini and large OCR models
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Quick answers
Is video-subtitle-extractor open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does video-subtitle-extractor need a GPU?
No. It runs on a CPU and can use a GPU when you have one.
Also in Data and scraping for AI
See all 50| Rank | Project | Score |
|---|---|---|
| 1 | doclingPython library that converts PDFs, Office files and more into structured text | 92 out of 100 |
| 2 | mineruParses PDFs, images and Office files into Markdown and structured JSON | 91 out of 100 |
| 3 | opendataloader-pdfPDF parser that outputs Markdown, JSON with bounding boxes, and Tagged PDF | 90 out of 100 |
| #4 | agent-reachCLI that gives coding agents read access to login-gated and Chinese web platforms | 88 out of 100 |
| #5 | markitdownPython library and CLI that converts office files, PDFs and audio to Markdown | 87 out of 100 |