Colorize photo — colourise black & white, automatically

Skies turn blue, skin looks lifelike, greenery comes back: free, no sign-up or watermark.

One photo at a time here. Need a batch, a different mode, or the before/after slider? Open the full tool.

Free, with nothing to install

There are no tiers to compare and no signup wall before you see the result: upload, and the colour comes back automatically. There is no app to download either; it is a page in your browser, and the result downloads once it is done, without watermark. The free tier allows ten runs a day, reset at midnight, so you can colorize photo online free for a whole afternoon's worth of prints. The before/after slider lets you judge the colours on your own upload rather than on a stock example.

How automatic colourisation works

A black & white photo still holds all its light and shade: what's missing is the colour. PicReviver uses a colourisation model trained on millions of colour photographs to predict what colour each part of the scene most likely had, then paints it back over the original tones. Because it has learned how the world usually looks, it gets skies, water, foliage, skin and everyday objects convincingly right, rather than adding flat or random tints. That is what an AI colorize photo tool does in general: predict plausible colour from tone and context rather than a human hand-tinting each region by eye, with the result back in seconds.

How to colourise a photo

Perfect for

Colourising suits old family portraits and vintage black & white prints, historical and archive photographs, and any monochrome scan you would love to see in colour. If the picture has faded rather than simply lost its colour, restore it first and then colourise the cleaned-up result; if it's small, upscale it larger too. To colorize black white photo prints from a whole box, the free tier handles ten runs a day of up to five photos each, and you can colorize photos straight from a flatbed scanner. A phone snapshot of an old black and white photo works the same way, uploaded straight from the camera roll, and so does a scanned negative.

What the model is guessing, and what it cannot know

Colourising is not restoration. A black-and-white negative never recorded colour, so nothing is being recovered: the model is predicting the most likely colours for the things it recognises. That prediction is remarkably good for anything with a natural range and unreliable for anything specific. It's usually convincing on skin, hair, sky, grass, wood, brick, water and most food, where the plausible range is narrow enough that a good guess is often right. It's frequently wrong on clothing: a dress that was emerald green and a dress that was burgundy can leave identical grey tones, and the model will pick whichever it has seen more of. Some things are basically unknowable — a military uniform's exact shade, a football kit, a flag, a company livery, a car's paint colour, a painted front door. If it matters, that colour needs a human who remembers, not a model. Mixed lighting is the awkward case: a photo lit by both a window and a tungsten lamp can come back with skin tones that shift across one face.

How to check a colourisation before you print it

Look at four things at 100% zoom, in this order. Start with edges: colour bleeding past an outline — hair colour leaking into the background, a jacket colour crossing onto a hand — is the most common artefact and the easiest to spot. Then skin: patchiness, or a face where one cheek is warmer than the other, means the model was unsure, and faces are also where a viewer notices wrongness instantly. Third, find your own control objects — something in the frame whose colour you know for certain, a familiar coat, a post box, a brand's packaging, a national flag — and if that known colour is wrong, treat every other colour in the photo as decoration. Finally, check consistency across a set: colourise several photos from the same day and compare, and if the same jumper comes out differently, you're looking at a guess, not a record.

The practical workflow: restore and sharpen first, then colourise, because cleaner edges give the colour model better-defined regions. Then ask someone who was there. A ninety-year-old who remembers that the coat was navy will beat any model, and the two together produce something worth framing.

Above all, keep the black-and-white original as the archive copy. A colourised print is an interpretation of a photograph, in the same way a colourised film is an interpretation of a film. It's lovely to look at and it should never be the only version that survives.

Practical questions

Upload a black and white photo and the colour comes back automatically, at no cost, whether it is a scan straight off a flatbed or a quick phone shot of a print. A colour photo that has merely faded is a different case; see the note on photo color further down.

The best input for AI colourisation has sharp edges and even lighting, since that is what the colour model reads best. There is no separate mode for recolouring just the background: the model colours the full scene in one pass, backdrop included.

Long before AI, people did this work with actual paint: hand-tinting a black-and-white print is the older version of the same idea. As a photo converter it turns a black-and-white file into a colour one; it does not touch the shapes or the detail.

This isn't a photo editor with brushes or layers. If you want a photo editor free of charge for manual touch-ups afterwards, open the result in a separate image editor for that part. Colour and clarity are two separate passes here, each with its own model, so this is not an AI enhancer for sharpness. Use the AI for colour here, and pair it with the sharpen tool if the print is soft as well.

Who's actually using this

Between 3 August and 2 October 2026 the colourise tool on this page ran 145 times, on top of colourising in the main tool on the homepage, and PicReviver had visitors from 150 countries, most of them from the United States, India, France, the UK and Germany. A black-and-white family photo means roughly the same thing everywhere: someone wants to see a relative's face the way it might have looked, not just in grey. That is also why the honesty above matters: a model guessing at a stranger's coat colour is doing the same uncertain job whether the photo came from Warsaw or Ohio, and it is wrong in the same ways in both places.

Colourising a scan versus a phone photo of a print

The colour model works from tone, and tone is exactly what a phone photo of a print gets wrong. Shooting a print under a ceiling lamp adds a warm gradient across the paper, and the model reads that gradient as information: one side of a face comes out ruddier than the other, and a white shirt picks up a cream tint that was never there. A flatbed scan lights the whole print evenly, so the greys mean what they say.

If a scanner isn't an option, shoot near a window with the lamp off, keep the phone parallel to the print, and check the result for an even tone across the paper border before you upload. Fixing lighting at the source costs a minute and saves you arguing with a colour cast afterwards.

You can colorize image files from a phone or a flatbed scan, as JPG, PNG or HEIC: upload a black-and-white print to colorize a photo in one pass, with the result ready to compare on the slider above. Warm or uneven tone usually comes from how the print was photographed, as explained above, not from the original scene. There is nothing to install: you color a photo in the browser tab you already have open, the colouring runs on our own server, and the photo is deleted right after processing and the result within 45 minutes. This is a photo AI free to use without an account.

The steps above turn a black and white photo to color automatically, using context rather than a manual paint tool. You convert black and white photo to color the same way for a scan or a phone photo, though a flatbed scan gives more even results. Going from black white to color photo takes one upload, and the model works the same on a faded print as on a crisp one. If a colour photo has faded rather than turned black-and-white, restore photo color with care: the restoration tool sharpens detail but keeps the faded colours, while colourisation replaces them with new colours predicted from the grey tones.

What is colorize meant to do on this page? It adds plausible colour to a black-and-white photo from context, using patterns learned from millions of colour photographs; what AI means here is a well-informed guess, not a record. You can try colorize free on your own photo, without creating an account. Can you colorize a black and white photo that is also faded or scratched? Yes: restore it first to rebuild detail, then run the colorize AI pass on the cleaned-up result. To colorize pictures in bulk, for instance colorize black and white picture scans from one album, drop up to five in at once with the full tool.

Before you ask

By Sam Ridder — I build and run PicReviver on my own hardware, on my own. Who I am.