Ideal inputs
Upload one image with readable text and specify the target language, exact replacement wording, text locations, and layout elements that must remain.
Create a translated-image concept by asking the edit to replace visible text while preserving layout, style, and surrounding imagery.
Upload one image with readable text and specify the target language, exact replacement wording, text locations, and layout elements that must remain.
Use a clear, high-resolution image where the text and its surrounding design are visible.
Name the target language and provide verified wording rather than asking the model to invent a translation.
Compare every translated word, line break, alignment, font-like treatment, and unaffected visual element with the source.
The image-edit workflow creates a translated visual concept based on the source and target-language direction.
OCR, translation, exact typography, line wrapping, and small text are not guaranteed. Use a human or translation tool to verify copy before publishing.
Not automatically. Supply verified target-language copy and proof every word because generated text and translation can be wrong.
No guarantee is made about typeface, line breaks, or alignment. Rebuild important text in an editable design file.