Content Aware Scale
Fill Any Frame, Crop Nothing
Content-aware fill & smart crop — free, private, no upload
See what content-aware scaling does
How it works
- 1 Upload your image — drop any PNG, JPEG, or WebP file (up to 25 MB / 4000×4000 px). Nothing is uploaded to a server; the file stays in your browser's memory.
- 2 Choose a mode — Fill Frame to expand your image to a wider or taller ratio, or Smart Crop to shrink it by removing the least important pixels. Pick a platform preset or enter a custom aspect ratio.
- 3 Download the result — processing takes 2–10 seconds via WebAssembly. Compare the before/after, then download in PNG, JPEG, or WebP.
Resize for a specific platform
FAQ
- What formats are supported?
- PNG, JPEG, and WebP up to 25 MB / 4000×4000 px.
- Is my image uploaded?
- No. Content Aware Scale processes images entirely within your browser using a WebAssembly binary compiled from Go. Your image is decoded into the browser's sandboxed memory, processed by the seam carving algorithm, and written back out as a file download. No network request containing image data is made at any point. You can verify this by opening your browser's Network tab and watching for requests while the tool runs.
- What is content-aware scaling?
- An algorithm that reshapes images by adding or removing the least important pixels. It powers two modes: Fill Frame (expand to a target ratio) and Smart Crop (shrink without cutting edges). Faces, text, and subjects stay intact.
- What's the difference between Fill Frame and Smart Crop?
- Fill Frame expands your image to fill a target aspect ratio by intelligently stretching low-energy areas like backgrounds, sky, and walls. Smart Crop shrinks the image to a target ratio by removing low-energy pixels instead of cutting edges. Use Fill Frame when the target is wider or taller than your source; use Smart Crop when you need to trim a dimension down.
- How fast is it?
- Typically 2 to 10 seconds depending on image dimensions and your device's CPU. Processing runs on a dedicated Web Worker thread via WebAssembly, so your browser stays responsive. Larger images (closer to the 4000×4000 pixel limit) take longer. On modern laptops and phones, most social-media-sized images process in under 5 seconds.
- Can I protect areas from removal?
- Yes. Face detection is enabled by default and automatically protects detected faces from being removed or distorted during processing. You can toggle it on or off in the controls. When enabled, the algorithm assigns higher energy to face regions, ensuring seams route around them rather than through them.