Two former TikTok employees have launched Superpose, an iOS camera app that uses generative AI to show you how to pose before you take the shot. Founders Melody Chu and Jing Liu built it around a simple pitch: point the camera at yourself or a friend, and the app generates four suggested poses you can then try to match in real life. TechCrunch AI reports the app launched in July and has already passed 22,000 downloads, with users generating more than 190,000 poses so far.
It also comes with $2.2 million in backing from Khosla Ventures, OVTR VC, and Meitu, the Chinese smartphone and selfie app maker. That last investor tells you a lot about where this category is headed.
What Superpose Actually Does
At its core, Superpose is a camera app with an AI layer on top. Here’s how it breaks down, according to TechCrunch AI:
- Snap a reference shot: You take a selfie or a photo of someone else. That image becomes the starting point for everything the AI suggests.
- Get four generated poses: The app produces four alternative poses based on your photo. Some are subtle, some are extravagant, and TechCrunch AI notes the results can occasionally look uncanny. Chu admits as much and says the team is working on it.
- Save or match: You can save the generated poses for later, or use the match-pose feature, which helps you line yourself up in the frame to recreate the suggestion. This is the part that turns a novelty into a coaching tool.
- Pay only if you want more: Everyone gets five free generations per day. Beyond that, a pack of five extra generations costs $2.99, and 20 generations run $9.99. No subscription, just top-ups.
Who’s Behind It
The founders’ backgrounds explain why this isn’t just another filter app. Chu held product roles at Meta, Nextdoor, Roblox, TikTok, and Slack. Liu was a founding engineer at a 3D face-scanning startup and later worked on image and video models at TikTok. That’s a product person paired with someone who has shipped real computer vision systems.
The origin story is refreshingly honest. “My husband just takes awful photos of me,” Chu told TechCrunch. “It was a true pain point in our marriage where I don’t understand how he gets me to look just so terrible all the time.” She figured computer vision and generative AI had matured enough to solve that, and set out to build it.
Where It Sits Against Google and Adobe
Superpose isn’t alone here. As detailed in TechCrunch AI, Google shipped Camera Coach on Pixel phones last year to guide users on framing and composition, and Adobe added an AI critique feature to its experimental camera app in July that suggests changes before you capture.
The difference is scope. Google and Adobe are coaching the photographer on composition. Superpose is coaching the subject on body language. Chu says the goal is to become “the best camera to take portrait photos by following directions from AI,” and she’s deliberately staying out of the AI background and fantasy image business.
“We want to lean into the core of memory capture versus, say, putting you in a fantastical place that you’ve never been,” Chu said. “I think there are so many apps that do that well already, but we really wanna focus on actually capturing the lived moment and experience.”
Why This Matters
What stands out here is the positioning. Most consumer AI photo apps compete on how much they can fake. Superpose is betting people still want real photos, just better ones. That’s a narrower market, but it’s also one the big platforms haven’t fully claimed, and it dodges the growing fatigue around synthetic imagery.
The Meitu investment is the tell. Meitu built a massive business on selfie enhancement in Asia. If it sees pose coaching as the next layer, expect the feature to show up in phone cameras, not just standalone apps.
The Caveats
- iOS only for now. No Android mentioned.
- The uncanny-pose problem is real and acknowledged by the company.
- 22,000 downloads is a decent start, not a breakout. The free tier is generous enough that conversion to paid packs will be the number to watch.
The startup says it’s working on more generation styles, personalization, and coaching that happens directly in the viewfinder rather than after the fact. That last piece would turn Superpose from a pose generator into an actual real-time camera assistant, which is where the category clearly wants to go. More details are in the original TechCrunch AI report.