🎨 FLUX 3 Max: 10-second video in 11 seconds of generation
On October 19, 2026, the fal platform, in collaboration with Black Forest Labs, released FLUX 3 Max — a variant of the FLUX 3 model refined through post-training, RL, and inference optimization. According to fal, a 10-second video is generated in 10.95 seconds: 22 times faster than Seedance 2.5 (277.80 s) and 11.5 times faster than the original FLUX 3 (126.44 s) with comparable quality. In the Artificial Analysis Video Arena voting (4868 votes), the model scored 1139 Elo — 50 points higher than FLUX 3 (1089) and on par with the top tier alongside MiniMax H3 (1140) and Dreamina Seedance 2.5 (1138). Through fal's unified API, text-to-video, image-to-video, and keyframes-to-video (up to 10 frames) are available: clips ranging from 5–20 seconds in 720p with upscaling to 1080p.
🌍 Generation speed has approached the duration of the clip itself, changing the economics of iterative work: teams can iterate through dozens of scene variations in minutes instead of waiting minutes for each render. The release also demonstrates that a multiple increase in speed and quality is achieved through post-training and inference optimization on top of a ready-made model — without retraining the base.
👤 FLUX 3 Max is already available on fal.ai in the playground and via API. For the first three weeks, until November 7, 2026, a 75% discount applies: 720p costs $0.0425 per second, or about $0.43 for a 10-second clip instead of the usual $1.70 — you can compare speed and quality with Seedance 2.5 and MiniMax H3. Note: the stated times are average values for 16:9 720p, and generation takes longer in high reasoning mode.
Source 1: https://www.prnewswire.com/news-releases/fal-launches-flux-3-max-delivering-1-tier-image-to-video-quality-at-industry-leading-speeds-302913033.html Source 2: https://fal.ai/models/fal-ai/flux-3-max