FLORA is the creative workspace for design teams: an AI canvas that puts every model on one canvas, from first concept to finished campaign. Teams from Pentagram to Lionsgate use it to turn a single concept into hundreds of production-grade assets.
The Challenge
FLORA's promise is every model on one canvas, so the canvas is only as good as the models on it and how reliably they run. A single workflow can move from an image model to a video model to a 3D model, and new ones launch every week. In the end, FLORA's customers care about two things: the quality of the output and how easy it is to produce.
“A lot of our customers don't actually care about the model. What they care about is that it works and looks great.”
Orpheas Kofinakos, ML Engineer, FLORA
The Solution
FLORA runs more than 140 image, video, 3D, and audio models through fal's API. Early access, which gives FLORA access to new models on fal's API before their public release, lets the team wire them into the canvas ahead of time, so they are ready for FLORA's users at launch. A designer can start a concept in one model, extend it into video with another, and turn it into a 3D asset without leaving the canvas, and FLORA's engineers spend their time on the canvas, workflows, and tools that creative teams pay for rather than on running models.
“We've built a lot of trust over the past year. We can rely on fal.”
Alec Jo, Head of Applied AI, FLORA
Over the 30 days to October 19, 99.7% of FLORA's requests to fal completed without a server-side error, across more than 970,000 generations.




