youcutai
AI ImageFree preview

Remove Subtitles from Screenshots

Hard-coded subtitles sit in a predictable band across the lower fifth of the frame, usually white or yellow text with a dark outline or a semi-transparent box. The Text Remover auto mode is tuned for that band: it detects the subtitle line, including two-line subtitles and burned-in translations, and inpaints the area so the scene continues naturally. Creators use it to grab clean stills from their own videos for thumbnails and social posts, educators use it to reuse frames from their lecture recordings, and localisation teams use it to prepare frames for new-language subtitles. Because subtitle text often overlaps faces, hands or moving objects, results are best when the frame is a wide shot or the text sits over a floor, table or background; where a face is fully covered the model has to invent features, so pick a nearby frame instead. For the semi-transparent black bar behind some subtitles, extend the manual mask to the whole bar. Use this only on video you produced or hold rights to; it is not a tool for removing broadcaster logos or watermarks from other people's footage.

  • Tuned for the lower-frame subtitle band, one or two lines
  • Clean stills from your own videos for thumbnails and posts
  • Extend the mask to cover semi-transparent subtitle boxes
  • Pick a frame where text does not cover a face
  • No signup required

from 1 credits · Uploads are deleted automatically after 24 hours.

After
Before
⇔
BeforeAfterDrag to compare

Legal notice: Only remove text from images you own or are authorized to edit. Do not use this tool to strip copyright notices, credits or watermarks from third-party content. Terms

Other variants

Modes

How it works

  1. Step 1

    Upload the image

    Screenshots, memes, scans, product photos or social posts.

  2. Step 2

    Confirm detected text

    Boxes appear around every text block. Untick any you want to keep, or switch to manual brush.

  3. Step 3

    Erase and download

    AI rebuilds the background under the text. Download PNG or JPG at original size.

Works great for

Why youcut.ai

Multilingual OCR

Detects Latin, CJK, Arabic and Cyrillic text, plus numbers and symbols.

Texture-preserving fill

Wood grain, fabric and gradients continue naturally under removed text.

Keep some, remove some

Untick detected boxes to keep the text you want.

Original resolution

No downscaling; output matches your input pixel for pixel outside the mask.

Free preview

Check the result before spending a credit.

youcut.ai vs Adobe Photoshop

Featureyoucut.aiAdobe Photoshop
Automatic text detection✓—
Time per image~5 secondsSeveral minutes with clone stamp
PriceFree preview, 1 credit HDMonthly subscription
Works on mobile✓—

Loved by creators and teams

“Recycled last year's campaign visuals by removing the old dates.”
Ivy C. · Social media manager
“Client scans had orange camera dates on every frame. Fixed 80 of them in an afternoon.”
Paul S. · Photographer
“I remove answers from worksheet photos to make blank versions.”
Rania K. · Teacher
“Old promo text on my own banners is gone in one click.”
Diego M. · E-commerce seller
“The fill under text is genuinely clean on gradients, which surprised me.”
Chloe B. · Designer
“Cleaned up lecture slide screenshots for my notes.”
Sam O. · Student

Frequently asked questions

Can it remove subtitles from a whole video?+

Not yet; the tool works on images and individual frames. Video subtitle removal is on the roadmap.

Does it handle yellow subtitles with black outlines?+

Yes, outline and shadow are detected as part of the text and removed together.

What about a translucent black bar behind the text?+

Switch to manual mode and paint the whole bar so the fill covers the darkened area too.

Ready to try Text Remover?

Free preview in seconds. HD download from 1 credits.

Try it free

Powered by: Bria Eraser · FLUX.1 Fill [pro]

Only remove text from images you own or are authorized to edit. Do not use this tool to strip copyright notices, credits or watermarks from third-party content.