Photo restoration

Upscale an old photo to higher resolution

A small scan or a tiny family picture gets blocky when you enlarge it. The bot sharpens the picture and rebuilds fine detail, so it holds up better on a larger screen. It adds detail by estimation, so you check the result before you rely on it.

Make it in Telegram 7 images free, then from 49 Stars

2 credits per image
Result in 1–2 minutes
No sign-up

In short: Upscale an old photo with AI when the scan is small. See what gets rebuilt, what gets guessed, and how to check it before printing. 7 images free.

Where: the VisuMate bot in Telegram, no app or sign-up.
Cost: 2 credits per image, 7 images free, then packs from 49 Telegram Stars.
Time: 1 to 2 minutes per image, rerun with a new instruction at the same price.
Credits never expire, there is no subscription and no auto-renew. You pay once for a pack inside Telegram with Stars.

When you need this

A scan from a flatbed is small and looks blocky when zoomed.
A photo came from an old messenger and is only a thumbnail.
You need a version that looks acceptable on a TV or a monitor.

How to do it

1
Use the button on this page to open the bot and press Start. The sharpening mode opens.
2
Send the smallest and the biggest copies you have, one at a time, as files. In the caption write: "enlarge, sharpen, keep faces unchanged".
3
Wait 1-2 minutes. Open the result at full size and zoom into faces and edges before you save it.

Try it on your own photo: 7 images are free, no sign-up.

Make it in Telegram

What upscaling means here

Upscaling a photo normally just stretches pixels, which makes soft edges. The bot works differently. It looks at the picture, sharpens it and draws in fine detail such as hair, brick or fabric weave where it expects them. The framing stays as in the original. The result is a cleaner and clearer image, and it is still a guess at detail that the scan never recorded.

When it works and when it does not

Moderately small scans with clear shapes respond well. Tiny originals, pictures with heavy motion blur and faces only a few pixels wide are the hard cases. In those the AI invents features, and a face can drift away from the real person. For faces you care about, compare the result with another photo of the same person. A group shot with many small heads needs a check on every one of them.

Before you print

The bot does not promise a fixed pixel size and does not print. It returns the image at a good resolution for screens and small prints. For a large poster or a frame, open the file at full size and look for smeared skin, odd teeth and repeated textures. If the picture also has cracks or stains, repair those first with the restore mode, then sharpen.

Tips

A bigger original always beats a smaller one, so rescan at a higher setting if you still have the print.
Crop out empty borders before sending, so the effort goes to the subject.
Do not chain several upscale runs. One pass is enough to judge whether the picture responds.
Keep the original file next to the new one.

Questions and answers

How many times bigger can I make a photo?

The bot does not promise a fixed size or multiplier. It returns the image at a good resolution for screen and small prints, so check the result at full size.

Is the extra detail real?

It is the AI's estimate. It looks natural but it is not recovered from the original.

Should I send a file or a photo?

Send a file (document), so Telegram does not compress the picture before the bot sees it.

Can I upscale a photo I took with a phone from a print?

Yes. Shoot in daylight without flash or glare, then send the phone picture.

What does it cost?

2 credits per image. New users get 15 credits, which is 7 images.

What the bot does not do

Does not print photos and does not retouch passport or ID photos.
Does not promise a fixed pixel size. Check the result at full size before printing.
Heavily damaged faces are reconstructed by AI and may not look like the person. Compare with another photo.
Does not make video or animate the photo.

More in "Photo restoration"

All pages in this section