Online AI Old Photo Restoration — Repair Faded & Damaged Photos
Repair scratches, fading, yellowing, stains, blur, and small damaged areas while preserving the people and character of the original photograph.
Restoration strength
What is AI old photo restoration?
CLROAI AI Old Photo Restoration uses a managed image-restoration workflow to repair scanned family photos, portraits, and archival images. It can reduce scratches, folds, dust, stains, fading, yellow color casts, visible grain, and small damaged areas while aiming to preserve identity, clothing, background, and the period character of the source.
Upload a JPG, PNG, or WEBP image up to 24MB, choose a restoration strength, and optionally request natural colorization. An account is required, and the current Star task price is shown before processing. Generative restoration may infer details in badly damaged areas, so compare the result with the original and keep an untouched source scan.
How to Use
Upload the Old Photo
Upload a flat, evenly lit scan or photograph in JPG, PNG, or WEBP format, up to 24MB.
Choose Restoration Settings
Select a light, detailed, or deep repair mode and enable natural colorization only when you want a color interpretation.
Compare and Download
Review the restored result against the original with the comparison slider, then download the version you want to keep.
Key Features
Scratch and Crease Repair
Reduce visible scratches, fold lines, tears, dust, and small surface defects.
Fading and Yellowing Correction
Recover tonal balance and contrast from faded, dark, or yellowed prints.
Natural Face Detail
Improve recognizable facial detail while instructing the model to preserve identity and age.
Optional Colorization
Add restrained, natural-looking color when requested or retain the source monochrome character.
Before-and-After Review
Drag the comparison slider to inspect repaired areas before downloading.
Account-Scoped Processing
Uploads and results are associated with the signed-in account and are not publicly listed.
Frequently Asked Questions
What damage can AI old photo restoration repair?
It can reduce many scratches, creases, tears, dust spots, stains, fading, yellowing, blur, grain, and small missing areas. Results depend on how much visual evidence remains in the source.
Will restoration change the person's face?
The prompt explicitly asks the model to preserve identity, age, expression, and facial proportions. Severely missing or blurred areas may still require inferred detail, so compare the result carefully with the original.
Can it colorize black-and-white photographs?
Yes. Enable natural colorization for a plausible color interpretation, or leave it off to preserve the original monochrome appearance.
Which restoration strength should I use?
Start with Light Repair for mild fading and dust, use Detailed Repair for blur and visible damage, and reserve Deep Repair for heavily scratched or torn photos.
Which file formats are supported?
You can upload JPG, PNG, and WEBP files up to 24MB and choose PNG, JPG, or WEBP in the export controls.
How much does one restoration cost?
Sign in to see the current Star task price, your balance, and available plans.
Is the result suitable for printing?
Review the restored image first. If you need a larger print, restore the damage before using the Photo Upscaler for the final output size.
How should I prepare the source photo?
A high-resolution scan is best. If using a phone, keep the photo flat, place the camera directly above it, use even light, and avoid glare or shadows.