The Workflow: Multi-Model Cinematic Chainer Turning

The Workflow: Multi-Model Cinematic Chainer

Turning a random phone snap into a 20s cinematic teaser

I wanted to see how far I could push that one cool blurry screenshot we always love. Instead of just rolling the dice with one prompt, I built a Multi-Model Cinematic Chainer to actually direct the shots.

How I built it: The Setup: Used Flux to expand the frame into a wide cinematic view. The Vibe: Forced GPT Image 2 to lock in those specific sunset rays and haze. The Action: Chained Kling for the tricky stuff—like the 180° turn and the close-up walk on the canal edge. The Look: Finished it off with a 35mm analog grade for that film grit.

Really happy with how the continuity held up across all 5 shots.

This is my submission for the MORHPIC WORKFLOWS CHALLENGE! @Morphic

Give it a spin yourself: You can clone the whole pipeline and try it with your own shots here: https://www.morphic.com/en/workflows/019dc694-a131-71f6-b2a5-2af8cdaf931e/multi-model-cinematic-chainer