FLUX 3 Max image-to-video: writing prompts as shot lists works best
OdinLovis · x · 2026-10-20
A practitioner shares hands-on prompting techniques for FLUX 3 Max's image-to-video mode with three case studies: describe only the action and sound (e.g. a character crying out in French with horse snorts and wheels on cobblestones), specify action beats while noting "one continuous shot, no cuts" for scenes like a train crossing as a bridge collapses, and write dialogue as a shot list with timestamps ([0-5 s] slow push-in, [5-7 s] cut to reaction) plus designated audio cues. The takeaway: structured, storyboard-style prompts give tighter control over video output.
More from Multimodal
- OpenMOSS releases Apache 2.0 MOSS-Image: 5B DiT with CJK text rendering and JSON layout control — crystal_alpine · 2026-10-20
- MOSS T2I + FLUX.2 Klein Edit Upscale Pipeline Hits 4K, Reddit User Shows Off — ZootAllures9111 · 2026-10-20
- Turn your kid's stuffed animal into an AI cartoon with Nano Banana + Dreamina — AIandDesign · 2026-10-20
- MiniMax H3 tops AA video leaderboard as FLUX 3 Max debuts at No.2 with half the price — ArtificialAnlys · 2026-10-20
- Realtime Local AI Video Restyling: 720p 60fps on a Single RTX 5070 Ti — Fit_Revenue1928 · 2026-10-20
- Houdini node builds metric-accurate humans and animates them via NVIDIA text-to-motion — Fit_Revenue1928 · 2026-10-20