Remove Text From an Image
Text on an image — a burned-in caption, a meme's top-and-bottom lines, a subtitle, a sign, or a label you no longer want — is really just an overlay hiding whatever was behind it. The challenge is that letters have hard edges and often sit over detailed backgrounds. WatermarkRemoverAI inpaints the text region and reconstructs the scene behind it, so you get clean background rather than a smudged rectangle where the words were.
How to remove text from an image
Upload the photo
Add the image containing the unwanted text — a caption, meme line, subtitle strip or on-image label.
Brush over the words
Paint across the full text, including the space between letters and any drop shadow or outline. Covering the shadow matters — a missed halo gives the removal away.
Download the clean image
Export the photo with the background rebuilt where the text used to be.
Good to know
Always brush the text's outline/drop shadow too, or a faint halo remains. Text over plain backgrounds clears best; text across a face or fine detail is the hardest.
Frequently Asked Questions
Can it remove text over a busy or detailed background?
It reconstructs the background from surrounding pixels, so text over a smooth or repeating area (sky, wall, grass) disappears cleanly. Text sitting across a highly detailed subject — a face, fine patterning — is harder because there's more structure to rebuild convincingly.
Do I need to mask the text's shadow or outline too?
Yes. Meme and caption text usually has a dark outline or drop shadow. If you only brush the letters and leave the halo, the result shows a faint outline. Cover the whole glyph including its shadow.
Can I remove some words and keep others?
Yes — you only remove what you brush. Mask just the line or label you want gone and any other text on the image is left exactly as it is.
Will it recover the object hidden completely behind the text?
It plausibly reconstructs the background, not a specific hidden object. If large text fully covered something unique, the AI fills the area to match its surroundings rather than inventing the exact thing that was there.
Does this handle subtitles burned into a video frame?
For a single frame, yes — upload it as an image. To remove hard-coded subtitles from a whole video, use the video remover so the subtitle band is cleaned across every frame.
Related Guides
Automate it with the API
Send a file, get a job id back, poll for the clean result. Same AI engine as the web tool, in two requests.
curl -X POST https://api.watermarkremoverai.com/v1/watermark/remove-image/ \
-H "Authorization: YOUR_API_TOKEN" \
-F "files=@photo.jpg" \
-F "auto_detect=true"
# => {"uuid": "..."}
# Poll for the result
curl -X POST https://api.watermarkremoverai.com/v1/watermark/results/ \
-d "uuid=..."
import requests
API = "https://api.watermarkremoverai.com/v1/watermark"
headers = {"Authorization": "YOUR_API_TOKEN"}
with open("photo.jpg", "rb") as f:
job = requests.post(API + "/remove-image/",
headers=headers,
files={"files": f},
data={"auto_detect": "true"}).json()
result = requests.post(API + "/results/",
data={"uuid": job["uuid"]}).json()
print(result)
Prefer a manual mask? Send mask=@mask.png instead of auto_detect, or pass mask_x / mask_y / mask_width / mask_height (add mask_is_percentage=true for percentages). Optional dilation=5 grows the masked area.
Ready to clean it up?
Upload, mark the area, and download the clean result. No account needed to try it free.
Open the Tool