Make an image meet exactly what the destination demands.

Why exact-size image tools get it wrong

Most online image tools give you a quality slider and let you guess. You drop the slider to 60, save, check the file size, upload, get rejected, drop it to 40, save, check again. Ten minutes later you have a picture that looks like a fax and a deadline you have already missed.

Autonom Image Fit flips the workflow. You tell it the constraint first — “under 200 KB”, “exactly 1080 × 1080”, “US Passport” — and it works backwards. For a file size target, it binary-searches the JPEG or WebP quality to find the highest value that still comes in under your limit. If quality alone cannot get there, it downscales by 15% and tries again. Whatever it took, the result is a file that fits, and you can see exactly what it did.

What it does

Three ways to fill empty space when the shapes do not match

When your image and the target have different aspect ratios, one of three things must happen: pixels get cropped, empty space appears, or the image is distorted. Autonom never distorts. For the “empty space” case, you have three fill styles — and each one is right for a different kind of image.

• Blurred backdrop. The empty space is filled with a blurred, enlarged copy of the image itself. This is what Instagram does with non-square photos and what YouTube does with 4:3 video in a 16:9 player. Adjustable strength.
• Edge extend. Autonom samples the outermost pixels of your image and stretches them outward to fill the frame. Works best on images with soft or uniform edges — skies, walls, gradients.
• Solid colour. Fill the empty space with a colour you pick. This is what passport and ID portals expect — usually white or a specific background colour.

About upscaling — the honest version

If your source image is smaller than the target dimensions, Autonom will scale it up smoothly and notify you. This is genuine — the pixels really do exist in the output. What we cannot do is invent detail that was never in the original.

A 400 × 400 photo scaled to 1600 × 1600 is technically 1600 × 1600, but it will look soft compared to a photo originally shot at that size. This is a limitation of photography, not of software. If a portal demands a large ID photo, the only way to get sharp results is to start with a large source photo.

ID and government photos — grouped by format

Most of the world’s ID photo specifications reduce to the same two or three physical sizes. Instead of a flat list of countries, Autonom groups them by their actual photo format — because what matters is the pixel dimensions, not which country issued the requirement.

• 2 × 2 inch (51 × 51 mm) — 600 × 600 px @ 300 dpi. US Passport, US Visa, US Green Card, India Passport, India Visa.
• 35 × 45 mm — 413 × 531 px @ 300 dpi. UK Passport, UK Driving Licence, Schengen Visa, EU Passport / ID, Australia Passport, Japan Passport.
• 33 × 48 mm — 390 × 567 px @ 300 dpi. China Visa, China Passport.
• 50 × 70 mm — 591 × 827 px @ 300 dpi. Canada Passport.
• Stamp size (20 × 25 mm) — 236 × 295 px @ 300 dpi. Common in India for forms and applications where a smaller photo is requested.

Each preset applies its own maximum file size where the portal enforces one — 240 KB for US documents, 300 KB for India Passport, 1 MB for India Visa. Everything is produced at 300 dpi equivalent, which is what these portals expect.

Use cases this was built for

• Government and visa portals — “photo must be under 100 KB” is a hard rule. Autonom lands you under it without quality collapse.
• Job applications — many HR systems reject profile photos above a certain size without explaining why.
• University uploads — thesis portals and student systems often cap images at seemingly arbitrary byte counts.
• Marketplace listings — Etsy, Amazon, and others have size and dimension rules that trip up first-time sellers.
• Website and social assets — Open Graph cards, article heroes, and email headers need exact dimensions and file-size limits.
• Client deliveries — when you cannot upload that NDA-protected product shot to a random online compressor.

How to use it in 4 steps

1. Drop an image — drag it, paste it from your clipboard, or click to browse. JPG, PNG, WebP, GIF, BMP, or SVG.
2. Pick the constraint — under a file size, exact dimensions, a platform preset, or an ID / document preset.
3. Adjust fit or fill — Cover crops (choose the focal point with the green rectangle); Contain keeps everything and fills with blur, edge extend, or a solid colour.
4. Download the result — one click. The result panel shows the original and output side by side, with dimensions, file size, quality used, and the change percentage.

Frequently Asked Questions