When Google Labs quietly dropped Doppl on June 26, it wasn’t with the fanfare of a flagship Pixel launch, but with the soft click of a “Try it now” button on the App Store and Google Play. Doppl, an experimental AI-powered styling app, promises to solve the millennial shopping conundrum: “Will this look good on me?” by putting the answer directly in your pocket—no fitting room required.

At the heart of Doppl’s origins lies a feature first teased at Google I/O 2025: virtual try-on in AI Mode for Search. There, Google showed off a proof-of-concept that let you upload a snapshot of yourself and overlay garments from its Shopping Graph—50 billion items strong—onto your own body. Internally, the idea germinated in Google Search Labs. But with the initial feedback proving overwhelmingly positive, Mountain View executives green-lit Doppl as its own downloadable experience instead of a buried toggle in Search.

Using Doppl couldn’t be simpler:

1. Upload a full-length photo of yourself. The AI maps your body shape, posture, and silhouette in seconds.
2. Import any outfit image. Scroll through your camera roll, screenshot that thrift-store find, or grab a shot from social media. Doppl’s AI parses the garment—buttons, folds, drape—and tacks it onto your digital avatar.
3. Watch it move. Unlike static “virtual dressing rooms,” Doppl can generate short videos, showing how sleeves bend or skirts sway—offering a more convincing preview of fit and flow.
4. Save or share. Snapshots or clips can be exported for feedback—text your friends, post to your ‘Gram, or just stash them for later.

By decoupling try-on from any particular retailer, Google sidesteps supply constraints: you’re not limited to listings in the Google Shopping ecosystem. If you’ve got a pic of Grandma’s vintage blouse or a street-market find in Paris, Doppl can handle it. And if an uploaded snap is missing shoes or pants? Doppl’s “Smart Completions” can intelligently propose missing elements, filling in gaps so you get a full-look preview.

This isn’t Google’s first foray into AI wardrobe wizardry. In 2023, the company rolled out a feature within Google Shopping that let you see clothing on a gallery of diverse, pre-modeled body types—an attempt to address inclusion in e-commerce imagery. Doppl takes that further, making the experience hyper-personal: it’s not a model that looks like you, it’s you.

Under the hood, Doppl leverages a custom, fashion-focused image-generation model developed by Principal Scientist Ira Kemelmacher-Shlizerman and her team. This network was trained on thousands of garment-body interactions, learning how fabrics stretch over shoulders, bunch at elbows, or flare at hemlines—knowledge that a generic “image synth” simply doesn’t possess.

While Doppl is fun, it also raises questions about personal data. Google insists that all body scans and style experiments remain on-device by default, with optional cloud backup secured by the user’s Google account. Users must be 18 or older and hold a U.S. Google account to try Doppl today; there’s no word yet on an international rollout or a younger-audience version.

For now, Doppl runs on standard smartphones—no Lidar needed—but performance does vary. Older devices may take a few extra seconds to render videos, and extremely complex patterns (think intricate prints or reflective materials) can occasionally glitch. Google says it’s working on optimizations and broader fabric recognition.

Doppl is available now on iOS and Android in the United States.