Compress to a specific size
Set a size limit and adjust quality or dimensions.
Choose a tool. Send its result directly to the next step without exporting and selecting it again.
Image to Text (OCR)
Recognize English and Chinese text locally.
Open tool →Format conversion
Convert static PNG, JPEG and WebP images.
Open tool →Compress to a specific size
Set a size limit and adjust quality or dimensions.
Open tool →Crop
Move and resize the crop box or enter pixel values.
Open tool →Remove EXIF / GPS
Remove original image metadata by re-encoding.
Open tool →Choose an image, adjust settings and process. Exportable results appear after successful processing.
Images are processed in a local Worker, never uploaded or saved automatically. SVG, GIF, HEIC/HEIF, AVIF and animated images are not supported. Metadata removal does not remove visible text, faces or watermarks; review the image before sharing.
How to use & FAQ
Choose an image and task, adjust the format, quality or crop, then process. Check the result size and dimensions before exporting. Changing settings clears old results. Cancellation terminates the Worker and never changes the source file.
Compress to a specific size means at most the target, not an exact byte count. Content, format and browser encoders affect size; unmet targets are clearly reported with the actual result. Resizing/re-encoding can reduce quality; colors and orientation follow browser decoding.
Support depends on device memory and browser OffscreenCanvas/createImageBitmap capabilities. Unsupported output formats are disabled or reported explicitly; PNG is never mislabeled as WebP.
Use Ctrl+V / ⌘V to paste a screenshot while this page is focused.
Some pages produced no recognized text.
One image, up to 20 MiB, 16 million pixels and 8,000 pixels per side. Only the first frame of animated images is processed.
OCR starts only when you select Recognize. Engine and language models load from this site (English model ≈3 MiB; Chinese ≈2 MiB, plus the engine). Model files may be cached in this browser. Your images, PDFs and recognized text are never uploaded or saved automatically.
How to use
Choose an image and its language, then recognize. You can cancel while resources load or during recognition. Copy the result or export UTF-8 TXT.
Recognition quality
Use sharp, upright images with readable printed text. OCR may miss or misread characters; check names, numbers and punctuation. Handwriting and complex tables are not guaranteed. This is a local OCR tool, not an AI document analysis service.
Images pass between tools in this page’s memory. Refreshing, changing language or leaving clears the files. The next tool’s size and format limits still apply.
How to use
Choose an image and set the maximum size in KB. Allow resizing if needed, then compare the actual output size before exporting.