Image work, from compression to masking in one place.

Identify the need first, then pick a tool

Most people need one of four things: a smaller file (image compressor), a compatible format (format converter), a specific size (crop or bulk resize), or content hidden (mosaic). Deciding whether it is a size, format, dimension or privacy problem saves a lot of trial and error. If the goal is just to lighten a website or chat attachment, compression alone usually does it.

Compression: balancing quality and size

JPEG quality between 75% and 85% is visually indistinguishable for most content while cutting file size by more than half. Social platforms re-compress anyway, so there is no point keeping the original pristine. Converting PNG screenshots to WebP often saves over 60%, especially for large flat areas. When a platform demands "under 2 MB", target a size in KB instead of guessing quality levels.

Formats: when to keep PNG and when to switch to WebP

Need transparency? Keep PNG or WebP and never convert to JPEG, which fills transparent areas with white. Photos are smallest as JPEG or WebP; icons and illustrations belong in SVG rather than a bitmap. HEIC is Apple’s default and poorly supported elsewhere, so convert it to JPEG before sharing or uploading. AVIF compresses best but compatibility is still catching up — keep a WebP fallback for anything public.

Dimensions and framing: ID photos, avatars, grids

Common ID photo sizes are 295×413 px (1-inch) and 413×579 px (2-inch); many application systems enforce both pixel size and file size, so double-check after cropping. Avatars are usually cropped to a circle, so keep the face centred with a margin or half of it disappears. For long images on social feeds, split into a 3×3 grid and post in order; leaving even margins keeps the seams from showing.

Privacy: masking and metadata matter more than you think

Before posting an original photo, check two things: whether the frame contains documents, door numbers or shipping labels, and whether the file carries GPS and camera details (EXIF). Sharing the raw file publishes where it was taken, so strip EXIF first. Masks must cover the edges — a thin smear can often be recovered, and when sensitive data is involved it is better to cover too much than too little. All of this runs locally, which is why we refuse to upload your images.

Bulk jobs and fine tuning

With dozens of files, prefer bulk resize and bulk compression over doing them one by one. GIFs are heavy — consider extracting frames or switching to video, and SVGs usually carry editor cruft that halves on optimisation. When building a brand palette, extract the dominant colours from existing artwork rather than guessing; when you need the words inside a screenshot, OCR saves retyping them.

FAQ