Problem
Animated stickers are the most expressive thing in a group chat, and almost nobody can make one. WhatsApp's format is strict — 512×512 WebP, under 500KB, looping cleanly — and every tool that gets you there assumes you already know ffmpeg. So the people with the funniest photos are never the people who can turn them into stickers.
Why I built it
Because the hard part was never the animation model. Models that move a still frame are available to anyone. The hard part is the twenty minutes of encoding, resizing, and format debugging between a good idea and something you can actually send. That gap is the whole product, and it is small enough for one person to close.
Solution
Stika takes one photo, animates it in a style you pick — Ghibli, 70s black-and-white, pixel art, meme — and encodes the result straight into a valid WhatsApp sticker. You upload a frame and get something you can send. Everything between those two steps is hidden.
Stickers go to WhatsApp directly from the web app, get published to a public market, and can be liked by anyone without signing in.
Who it is for
Three groups, and they behave differently. Couples making private in-jokes. Friend groups building a shared sticker vocabulary. Meme communities animating a moment while it is still funny — the only one of the three where speed is the feature.
Outcome
- Photo-only flow after text-to-video was cut for being the weaker path.
- Generation runs in the background, so the tab is never blocked on a slow model.
- Free tier capped at three generations per user, reserved before the paid model call so a failed job never costs a paid credit.
Learnings
The expensive part of a generative product is not the model, it is deciding what to remove. Cutting caption text and text-to-video made the remaining flow obvious.