Use case
Colorize a black and white photo
Add color to a black and white photo without hand-painting a single pixel. MightyPhoto infers skin, fabric, and scenery from the grayscale detail, and if you remember a real color, you can tell it.
Step by step
How to do it
Upload the black and white original
A scan or a phone photo of a print both work. If the photo is also damaged, restore it first and colorize the clean version: the AI reads color cues from detail, and scratches confuse it.
Pick the era
Leave it on Auto-detect, or choose 1900s–1930s, 1940s–1950s or 1960s–1970s to keep the palette plausible. If you know a real detail, add it in the optional box: “her dress was pale green” beats any guess.
Adjust and download
If a color lands wrong, name it in the optional box and run again. One credit per attempt, so dialing in a specific shade is cheap.
Which photos colorize well
The AI picks colors from what it can see: the tone of skin, the texture of wool against cotton, the brightness of a sky. A sharp photo with a full range from black to white gives it the most to read. A faded print with gray blacks and dull whites gives it little, and the colors come out muddy.
If the photo is scratched, creased or faded, run Old Photo Restore first and colorize the clean copy. Scratches and dust get colored along with everything else, and they are harder to remove afterwards. Restore and then colorize is two credits. For a photo in good condition, Colorize B&W on its own is enough.
Using the era and what you remember
Leave Era on Auto-detect for most photos. Choose a decade when you know it and the clothing or cars matter: 1900s–1930s gives a muted, early-color palette, 1940s–1950s and 1960s–1970s keep fabrics, paint and cars plausible for the period.
The Anything else? box is where colorization gets accurate. The AI cannot know that your grandfather’s car was dark green or that your mother’s wedding flowers were yellow. You might. Type it in, up to 200 characters: “her dress was pale blue, his uniform was olive green”. Ask relatives before you start. One remembered detail fixes more than any setting.
If one color lands wrong, name it and run again. Each attempt is one credit and keeps the earlier versions, so you can compare them.
Colors are an interpretation
Colorizing adds information that was never in the photo. Skin, sky, grass and common fabrics come out convincing. Specific items, like a dress, a flag or a painted door, are educated guesses unless you supplied the real color.
That matters if the photo goes into a family history, a genealogy site or a local archive. Keep the black-and-white original with the colorized copy, and say it was colorized when you share it. An artist who hand-colorizes a historical photo can spend hours researching uniforms, paint and dyes. The AI version takes about ten seconds and works best when you give it the facts.
In the tool
Popular settings
Pick these in the tool. No prompt to write.
Auto-detect
Right for most photos.
1940s–1950s
Naming the decade keeps clothing and cars period-correct.
1960s–1970s · “the car was dark green”
Real details you remember beat anything the model can infer.
1900s–1930s
Muted, early-color palette for the oldest portraits.
Questions about colorize black and white photos with ai
How accurate are the colors?
The AI infers plausible color from brightness, texture, and context: skin, sky, foliage, and common fabrics land reliably. It cannot know that a particular dress was burgundy rather than navy. If you know a real detail, add it in the optional box and it will be used.
Can I colorize and restore in the same step?
Yes. Old Photo Restore has an “Also colorize” option. For a badly damaged photo, two passes (restore, then colorize the clean copy) usually give a better result. Each pass costs one credit.
What about a color photo that has faded?
Use Old Photo Restore and tick Fading. It corrects the faded color instead of inventing a new palette.
How much does it cost?
One credit per photo at HD, two at 2K. Credit packs start at $4.99 for 10 credits.
