Live tool · Aliyun Qwen
Select multiple text regions and rewrite them in one run while keeping style. Each submit costs 1 point(s).
No replacements yet. Upload an image and click a green text box.
Click multiple boxes, fill in new text for each, then start rewrite.
Change “” to:
Guide
MP Picture Text Style detects text in an image with OCR, lets you pick one or many regions, and rewrites those words while aiming to keep font look, color, and perspective.
Unlike editing a separate text layer in a design file, many photos bake text into pixels—signs, packaging, posters, UI mockups. Text Style uses OCR to locate readable strings, then an image-edit model replaces the glyphs so the surrounding art stays intact as much as possible.
Typical workflow:
Text Style is useful when you need localized or corrected wording without a full redesign:
Detection quality drives everything else:
Each successful rewrite costs points regardless of how many regions you batch in one submit. Upload common image formats under a practical size limit. OCR accuracy varies by language, lighting, and font; always review the result before publishing.
Erase unwanted people or clutter first, then rewrite remaining text. Outpaint if you need a different canvas size for the final graphic. Colorize older prints before adjusting captions.
FAQ
Upload an image for OCR, click green boxes to add replacements, enter new text, then submit. AI tries to keep the original style.
Yes. Add several replacements, then submit once. Each successful run costs 1 point(s).
Make sure text is clear with good contrast; try Re-detect text. Handwriting or heavily warped fonts may fail.
OCR works best on clear printed text. Support varies by script and quality; always verify detected boxes before rewriting.
Yes. Batching multiple replacements in one run still costs the standard Text Style point fee once per successful submit.