fal ships FLUX 3 Max, a video model 22 times faster than Seedance 2.5
Post-trained by fal Research with Black Forest Labs, the model generates a 10-second 720p clip in about 12 seconds and is 75% off through 7 November.
fal released FLUX 3 Max on 19 October, a text-to-video and image-to-video model post-trained by fal Research in collaboration with Black Forest Labs on top of the FLUX 3 base model. It generates a 10-second, 720p clip in about 12.45 seconds, roughly 10 times faster than the original FLUX 3 and 22 times faster than Seedance 2.5 at comparable quality, which fal says takes 277.80 seconds for the same job.
On Artificial Analysis’s preliminary image-to-video arena, FLUX 3 Max scores 1,139 Elo, a tie for first place and 50 points above the base FLUX 3 model. It’s billed per output second, at $0.0425 a second through 7 November, 75% off the $0.17 list price, with no subscription or minimum.
Why it matters: video generation has mostly meant long waits or expensive batch jobs. A model fast enough to return a 10-second clip in roughly the time it takes to read this sentence changes what’s viable to build interactively, previews, variations, iteration loops, rather than queue-and-wait pipelines.
The caveat: the Elo score and speed comparisons are from fal’s own benchmarking against an early, preliminary leaderboard, not an independent third-party test.