Black and white photos come to life in natural color. The AI paints only the color layer, so your photo keeps every bit of its original sharpness, entirely on your device.
Somewhere in every family's archive is a face that has only ever existed in gray: a wedding portrait, a soldier's photo, a grandmother as a young woman. Color changes how those pictures feel; suddenly the person seems near enough to have been photographed last week. The colorization services that promise this either charge per photo or route your family's archive through their servers.
This tool does it on your machine. DDColor, a modern colorization model released openly by Alibaba's DAMO Academy, predicts natural colors for the whole scene: skin, sky, foliage, fabric, and stone each get treated as what they are. The model downloads once (about 54 MB), caches in your browser, and works offline afterward. Nothing you drop here ever uploads.
AI colors are plausible, not documentary. The model learned what the world usually looks like, so it will make grass green and skin warm, but it cannot know that a specific car was blue or a specific dress was red. For historical work, treat the output as an interpretation. For family photos, most people find the interpretation is exactly what they wanted: the person, present again.
Drop the photo here and the DDColor AI model, running in your browser, predicts natural colors for every part of the scene: skin, sky, foliage, clothing. Compare with the wipe slider, adjust the vibrance to taste, and download at full resolution. No account, no watermark, and no per-photo limit, because your own device does the work.
No. The AI model downloads to your browser once (about 54 MB, cached for later visits) and every colorization runs on your device. The photo never touches a server, which matters when the photo is the only copy of someone you love.
No, and this is the part most people do not expect: the AI predicts only the color layer (chrominance), while your photo's original brightness detail (luminance) is kept exactly as-is. Every edge, texture, and grain stays at full sharpness; only the color is new.
They are plausible, not documentary. The AI learned from millions of photos what things usually look like, so grass comes out green and skin comes out natural, but it cannot know that a particular dress was red. Treat the result as a lovely interpretation, and use the vibrance control to keep it subtle if you prefer.
Yes. Drop a faded or color-cast photo and the AI re-colors it from the brightness alone, replacing the washed-out tones with fresh, natural color. It is a surprisingly effective revival for 1970s-90s prints that have shifted orange or magenta.
The same resolution you started with: the color layer is computed by the AI and blended into your original photo at its native size, and the download is PNG or JPEG with no watermark. On phones you can also save straight to the camera roll.