#1515th of 56 in Image and videoFor image, video and design, image and design
remove-ai-watermarks
CLI that strips visible AI marks, metadata and invisible watermarks from images and video
- Stars
- 6.3k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Last release
- Nov 2026
- Language
- Python
15 on 21 Apr 20266,348 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Python CLI and library that finds and removes AI provenance marks from your own images, video and text. It covers visible labels (Gemini sparkle, Veo, Sora, Doubao, Kling and others), C2PA/EXIF/XMP metadata, and invisible pixel watermarks via diffusion regeneration. Visible and metadata paths run without a GPU; invisible image removal needs CUDA.
Who it is for: Creators and researchers cleaning provenance marks from content they generated
Strengths
- Visible and metadata removal run on CPU with OpenCV, MI-GAN or LaMa fill backends
- Handles images, video (MP4, MOV, WebM, MKV) and text provenance from one CLI
- Video path scans all frames for a stable mark and copies audio without re-encoding
- Installable by feature extras, so metadata-only use pulls no torch
Weaknesses
- Invisible image removal is CUDA-only, with no CPU or MPS fallback
- No local SynthID pixel detector; detection relies on signed provenance
- qwen21-zimage profile is restricted to non-commercial use by the Qwen Research License
- Some jurisdictions, including China, restrict removing AI labels; legal risk falls on the user
What it needs
- GPU optional
- Needs ffmpeg, NVIDIA CUDA (invisible removal), PyTorch (diffusion extras)
- Models: Qwen-Image-2512 Lightning, SDXL, Chroma1, Z-Image, MI-GAN
Quick answers
Is remove-ai-watermarks open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does remove-ai-watermarks need a GPU?
No. It runs on a CPU and can use a GPU when you have one.
Also in Image and video
See all 56| Rank | Project | Score |
|---|---|---|
| 1 | frigateLocal NVR with realtime object detection for IP cameras | 92 out of 100 |
| 2 | MoneyPrinterTurboGenerates short videos from a topic, with script, footage, voiceover and subtitles | 88 out of 100 |
| 3 | toonflow-appInfinite-canvas tool for AI short drama and comic video production | 86 out of 100 |
| #4 | ComfyUINode-graph editor for running diffusion image, video, audio and 3D workflows | 85 out of 100 |
| #5 | InvokeAIWeb UI for diffusion image generation with inpainting canvas and node workflows | 85 out of 100 |