Free AI old photo restoration — repair damage, colourize, enhance faces
Restore and colourize old photos — free, AI-powered, no sign-in.
3 free guest generationsNo sign-in required10 free with Google accountNo credit card
Upload a damaged, faded or black-and-white photo and let AI restore it instantly. Repair scratches, tears, fading and discolouration — or colourize black-and-white family photos with historically accurate colours. 3 free guest generations per IP; sign in with Google for 10 one-time free generations.
Pick a restoration type:
3 free generations per IP — no sign-up needed. Optional: sign in for 10 one-time free generations and saved creations.
Up to 8 MiB, 4 megapixels and 4,096 px per side. Static images only. Use the photo editor to resize first.
A new direction for your lines.
Upload an image, describe your change, then click Generate image. No sign-in required for your guest allowance.
No image has been generated on this page yet.Keep the whole composition in view. Outlines, small details and proportions can change; this is not a CAD or construction tool.
What this AI photo restoration tool can fix
- Scratches and tears. Physical damage like scratches, creases, tears and fold marks appear as white or dark lines. AI fills these using surrounding pixel context, blending seamlessly with the original image.
- Fading and discolouration. Old prints fade to sepia or develop uneven colour shifts over decades. AI restoration recovers natural tonal range and can optionally reintroduce colour.
- Blurry or soft faces. Old photos often have blurry faces due to slow shutter speeds, low-resolution film or scan quality. Dedicated face restoration sharpens features and recovers detail while preserving the original likeness.
- Dust spots and foxing. Small brown or white spots caused by dust, moisture or paper degradation over time.
- Colourize black-and-white photos. Add natural, era-appropriate colours to black-and-white or sepia photos. Describe the decade, setting and any known colours in your prompt for the most accurate result.
- Water damage. Blotchy stains, bleached areas and water marks from moisture exposure or storage in damp conditions.
- Vintage and antique photos. Daguerreotypes, tin types and early 20th-century prints often have multiple damage types at once. Use the full restoration prompt for best results.
When to restore old photos
- Family memories and genealogy. Grandparent and great-grandparent photos are among the most common restoration requests. A restored and colourized photo makes a meaningful gift for family reunions, anniversaries and memorials.
- Wedding and milestone photos. Old wedding photos from the 1950s–1980s are often faded or damaged. Restoration preserves these irreplaceable images and colourization brings them into a modern context.
- Military and historical photos. Service portraits, wartime photos and historical documents restored and colourized to honour the people in them.
- Vintage photos for printing. Restore a digitized scan before printing as a large-format canvas or framed print. Higher resolution scans give the AI more detail to work with.
Tips for the best restoration results
- Scan at high resolution. Scan or photograph the original print at the highest resolution available before uploading. More pixels give the AI more original detail to preserve and restore.
- Describe the specific damage. "Remove a horizontal scratch across the lower half" is more precise than "fix damage". Specific instructions produce more targeted results.
- For colourization, describe the era and setting."1940s outdoor portrait, natural skin tones, blue summer dress, green lawn" gives the AI historical context for accurate colour choices.
- Use multiple passes for heavy damage. For severely damaged photos, restore structural damage first, then colourize in a second generation.
- Mention face count for group photos. "Restore all four faces in this group photo" helps the AI prioritise face detail.
Old photo restoration — common questions
Is AI old photo restoration free?
Yes. Guests get 3 free restorations per IP without signing in. Google accounts receive 10 one-time free generations — no credit card required. Valid paid credits are spent first when available.
Can AI colourize black-and-white photos?
Yes. Upload your black-and-white or sepia photo, use the "Colourize only" preset or describe the colours you want: "1950s family portrait, natural skin tones, blue dress, brown hair". The AI applies historically appropriate colours based on your description.
Can it restore blurry or soft faces?
Yes. Use the "Fix faces" preset or describe: "sharpen and restore the faces in this photo". The AI sharpens features and recovers detail while trying to preserve the original likeness. Best results come from photos where the face is at least partially visible.
Can AI restore severely damaged photos?
AI significantly improves heavily damaged photos, but severe damage means less original pixel data. The AI fills missing areas based on surrounding context, which may introduce minor reconstruction artefacts in extreme cases. For best results, describe what should be in the damaged area.
How do I prepare a photo for restoration?
Scan the original print at the highest available resolution (at least 600 DPI, ideally 1200 DPI) before uploading. More pixels give the AI more to work with. JPEG, PNG and WebP up to 8 MiB are supported.
Can I restore a photo of a grandparent or family member?
Yes — family portraits are the most common restoration request. Use the "Grandparent photo" preset which prompts for full damage repair, face enhancement and optional colourization in one generation.
Does AI photo restoration upload my photos to a server?
The local format and content check runs on your device before anything is uploaded. AI generation processes the image server-side under a reviewed generation workflow when you click Generate.
What is vintage photo restoration?
Vintage photo restoration refers to repairing photos from earlier eras — typically pre-1980s prints, daguerreotypes, tin types or early colour prints that have suffered physical damage, chemical fading or discolouration over decades of storage.