google's omni model + podcast ugc is actually insane
this image is the whole workflow
prompt
creator
synchronized
edited
final
one input turns into multiple podcast style ads with different creators, matched audio, captions and a finished export
i’m not stitching clips together
i’m not manually syncing voiceovers
i’m not sending anything to an editor
i drop in the angle, generate the creators, and the whole thing comes back ready to publish
i’ve tested this format to death and it’s one of the best converting styles i’ve touched because it feels native, holds attention and sells without looking like a normal ad
so f*ck it, i’m giving away the whole system
> the exact workflow from prompt to finished video
> how i generate podcast style ads with multiple creator variations
> how i sync audio, captions and final edits automatically
> the script structure that makes ai talking ads feel real
> my starting image rules for realism and better outputs
rt + comment "podcast" and i’ll send it
(follow for dm)