Help Center

Everything you need to know about removing watermarks with Unwatermark.

How It Works

Upload a file, and our AI pipeline automatically detects and removes watermarks.

Upload
AI Detection
Neural Inpainting
Download

Step-by-Step Guide

Step 1
Upload your file

Drag and drop a file onto the upload zone, or click to browse. Supports PNG, JPG, BMP, TIFF, WebP, PDF, and PPTX files up to 50 MB.

Step 2
AI processes your file

The detection pipeline automatically scans for watermarks using AI vision analysis. For multi-page files (PDF, PPTX), each page is processed individually with real-time progress updates.

Step 3
Download your clean file

For images, compare before and after with an interactive slider. For documents and presentations, download the cleaned file directly. Click Start Over to process another file.

Best Results

Unwatermark works best with certain types of watermarks. Here's what to expect.

Works well

  • Small corner watermarks on simple backgrounds
  • Stock photo text overlays (Shutterstock, Getty, etc.)
  • Copyright text and "DRAFT" stamps
  • Watermarks on solid or gradient backgrounds
  • Single-image files (PNG, JPG)

May have limitations

  • Small watermarks near dense content text
  • Large diagonal overlays covering the whole image
  • Tiled/repeating watermark patterns
  • Watermarks overlapping complex graphics or photos
  • Multi-page PPTX with varied slide layouts

Supported Formats

  • Images -- PNG, JPG, BMP, TIFF, WebP. Single image files up to 50 MB. Results include a before/after comparison slider.
  • PDF -- Multi-page documents (up to 15 pages). Each page is rendered to a high-resolution image, cleaned, and reassembled into a new PDF. Note: vector content is rasterized in this process.
  • PPTX -- PowerPoint presentations (up to 15 slides). Embedded slide images are processed in-place, preserving all slide layout and formatting.

FAQ

Is my file sent to the cloud?

Your file is sent to AI services for watermark detection (to locate the watermark). The actual removal uses neural inpainting via our processing pipeline. Files are not stored after processing -- they are discarded once you download the result.

Why does processing take a while?

The pipeline runs multiple detection passes and neural inpainting for each page. For multi-page files like PPTX, each slide is processed individually. Processing typically takes 15-60 seconds per page depending on complexity.

What if the watermark isn't fully removed?

Some watermarks, especially very small ones near dense content, are difficult to remove cleanly. The pipeline prioritizes preserving your content over aggressive removal -- it's better to leave a trace than to damage the surrounding text or graphics.

Does this work with NotebookLM watermarks?

Yes, NotebookLM watermarks are one of the primary use cases. The tool has specific detection logic for the small "NotebookLM" text that appears in the bottom-right corner of exported slides. Results vary depending on how close the watermark is to slide content.

Is this free to use?

Yes, completely free. No signup, no account, no usage limits. The tool is a portfolio project by CushLabs AI Services.