Lottie AI - Editable vector animations from a text prompt

by•
Motion generated as editable vector animation, built with builders in mind. Get a lightweight Lottie you can recolor, restyle and resize without regenerating, with custom controls for every animation. Add hover and click interactions with a simple prompt. Start with Turbo for quick ideas, or switch to Quality when you have a solid plan. No motion-design skills needed. Export to dotLottie, JSON, GIF, video and more.

Add a comment

Replies

Best

Hey Product Hunt 👋 I'm Ben, the product manager on Lottie AI.

We built it for people who need motion in their product, not a new animation career. Developers and builders working in Cursor, Claude or v0, and product designers who need a loader, an empty state or an onboarding moment without waiting for the next sprint.

You describe the animation in a sentence and you get a real Lottie back. Not a video, a vector file that stays sharp at any size and stays editable. You refine it by chatting, export it for web and apps, and open it in Lottie Creator when you want to keep going by hand.

Over the last few weeks we put it in the hands of designers and developers at our events, and a lot of what you see today comes from what they told us. It's free to try, so the best way to judge it is to throw your own idea at it.

What I'd love from you: what would you use it for first, and where did the result fall short of what you pictured? I'll be reading every comment today, and it all goes straight into what we build next 🙏

Hey Product Hunt 👋

I'm Robert from the Lottie AI engineering team.

Lottie AI turns a prompt into a real Lottie file: vector, tiny, every layer named, with sliders built for that exact animation.

Under the hood, Lottie AI doesn't push keyframes around. It writes code in our own animation language, and our compiler turns that code into Lottie.

The language already has bones and IK, physics, deformers, interactive state machines and 300+ built-in functions. When the AI makes a mistake, the compiler points at the exact spot and says how to fix it. The AI repairs its own code before you ever see it.

That's the bet: every feature we add to the language is something new Lottie AI can make, and more are on the way.

Try it, and show us what you made 👇

Hey Product Hunt 👋 I'm Naail, one of the engineers behind Lottie AI.

Most AI animation tools hand you a video. You can watch it, but you can't really change it. We wanted the opposite: describe an animation, get a real Lottie file, and then keep shaping it like a designer would.


Under the hood, the AI doesn't draw frames. It writes a small animation language that our compiler turns into clean, lightweight Lottie. That's why the result stays editable instead of being a one-shot render:

🎨 Controls: tweak colours, speed and the overall vibe without re-prompting
🧩 Objects: reorder, hide or remove any part of the animation
💬 Chat to refine: "make the bounce softer", "add a shadow", and it edits what's there instead of starting over
📦 Export: Lottie and dotLottie for web and apps, and other formats.

It's built for product designers and developers who need UI motion (loaders, empty states, onboarding, icons) and need to get it done quickly.

We're really looking forward to see what you build with it.

So proud of the team who worked on this. Watching it come together has been really cool.

Hey ProductHunt! Developer on the Lottie AI team here.

As a dev, motion was always the thing I’d skip. Not because I didn’t care, but because getting it on-brand meant a designer, a handoff and a few rounds of “can you make it a bit calmer?”

What changed for me: every animation comes with its own control panel, built for that specific animation. I generate once, then match our brand colours, tone the motion down and adjust timing myself, and it updates live. No new prompt, no credit, no waiting.

A few minutes of iterating and it looks like it belongs in our product. Curious what you’d build with it!

Hi Product Hunterssss,

Amirul from the engineering team and I’m excited to finally share what we’ve been building!

Our goal is simple: make it possible for anyone to create production-ready animations with AI, without spending hours manually keyframing. Just to highlight a few:

  • Start from your own artwork: import an SVG or Lottie and animate it, or drop in a raster image and get clean, editable vectors back

  • Make it interactive: add hover and click states so the animation responds to users

  • Fast or quality: fast mode for quick ideas, quality mode for detailed scenes

  • Versions and remix: go back to any earlier version, export to MP4 or HTML, or share it for others to remix

Behind the scenes, we’ve been tackling some interesting engineering challenges, from image understanding and segmentation to raster-to-vector conversion. What I’m most excited about is that the result is something you can actually work with, not just something that looks good.

There’s still plenty we want to improve. If you try it on a messy image, I’d love to see how it holds up

Love seeing this live. Making animation this easy for builders is such a good mission.

Made this on and I've been loving how i can just keep prompting tweaks and adjust it with the controls. No keyframes needed.

Finally, creating motion without spending hours fiddling with timelines and keyframes. Incredible work!

I can’t animate at all but I managed to create a pretty decent Harry Potter-inspired owl with Quality mode. Love this! Congrats team on the launch. 🎉

Congrats team on the launch!! 🚀

Being open about the code behind it goes a long way in building trust, especially with developers.

123
Next