Video Subtitle Remover is a free, open-source tool that uses AI to erase hard-coded subtitles and text watermarks from videos and images, then rebuilds the cleaned frames at the original resolution. The project has earned more than 13,300 stars on GitHub, is licensed under Apache-2.0, and was still being updated in 2026. Everything runs locally on your own machine โ no third-party API keys, uploads or per-minute fees.
Hard subtitles burned into a video cannot simply be switched off, and watermark text is deliberately hard to remove. Video Subtitle Remover tackles both with deep-learning inpainting: it detects the subtitle or watermark regions frame by frame and reconstructs what should be behind them, producing a clean video or image without shrinking the resolution. It is a popular choice for re-editing downloaded clips, cleaning footage for reuse, and preparing subtitle-free masters before adding fresh, properly timed captions in another language.
The workflow is deliberately simple: point the tool at a file, let it locate the text areas automatically (or guide it to the subtitle zone), and export the cleaned result. Because processing happens on your own GPU or CPU, long videos stay private and there is no queue, watermark or resolution cap of the kind online removal services impose. Batch-friendly local processing makes it practical for cleaning whole seasons of episodes or entire folders of images in one sitting.
Video Subtitle Remover is completely free. Download it from GitHub and run it on your own hardware; there are no subscriptions, credits or paid tiers.
The tool is ideal for video editors, translators, content creators and archivists who work with footage that has burned-in subtitles or text watermarks โ for example when localizing a video, cleaning stock clips, or restoring text-free versions of short-form content.