Which AI image or video model for which shot
Pick the model by the job. For images: Nano Banana for studio direction, edits and legible text; GPT Image for typography, posters and layouts; Midjourney for stylised, art-directed images; FLUX for photorealism and long, detailed prompts; Seedream for editing and combining several reference images. For video: Seedance for camera moves and multi-shot clips; Kling for clips up to 15 seconds with native audio.
Image models
Nano Banana for studio direction, edits and legible text
Name the light and the lens; that gets closer than a list of adjectives.
54 Nano Banana prompts with results →GPT Image for typography, posters and layouts
Put the exact words in quotes and say where they sit on the image.
49 GPT Image prompts with results →Midjourney for stylised, art-directed images
Keep it short and evocative: medium, style and mood words, plus its own parameters.
22 Midjourney prompts with results →FLUX for photorealism and long, detailed prompts
Describe the scene in full sentences (subject, setting, light, lens), not a pile of tags.
12 FLUX prompts with results →Seedream for editing and combining several reference images
Describe props precisely; a loosely described prop drifts from one take to the next.
8 Seedream prompts with results →
Video models
Seedance for camera moves and multi-shot clips
Write it like a shot list: one camera move per beat, the light, then what changes.
83 Seedance prompts with results →Kling for clips up to 15 seconds with native audio
One continuous action and one camera move read best.
8 Kling prompts with results →