CapyImagine · tool no. 2
.
Random image & video prompts, tuned per engine. All local.
CapyImagine rolls random, ready-to-paste prompts for Midjourney, Flux, SDXL, Gemini and video models, written in each engine's own syntax, free and with no API key.
Updated · by Unfolding Dimensions
01 · Scene controls
02 · The prompt
Gemini · seed —…
Gemini → No negatives field — exclusions ride inline, so they stay short.
Every control re-words this scene; Randomize draws a new one.
The guide
How to generate random AI image and video prompts
- Step 1
Pick your engine — Gemini, Midjourney, Flux or SDXL for images; Kling, Runway or Seedance for video.
- Step 2
Roll a prompt. Subject, setting, lighting, lens, medium and composition are drawn at random and written in that engine's own syntax.
- Step 3
Copy it into your engine, or roll again until one sparks.
Every image model reads prompts differently. Midjourney wants its parameters as flags, SDXL takes a separate negative prompt, and video models want camera motion. CapyImagine writes each prompt in the dialect of the engine you picked, aspect ratio and negative clauses included.
It is a cure for the blank prompt box: a random but well-formed starting point you can edit, rather than a finished artwork.
Questions
Does it use AI to write the prompts?
No. Prompts are assembled in your browser from curated lists, so generating one is instant, free and needs no API key.
Will the prompt get blocked for naming an artist?
For engines that refuse prompts naming living artists — Gemini and the video models — those references are left out, checked against a list of living artists. Engines without that rule can get them.
Does it generate the image too?
No. It writes the prompt; you paste it into the image or video tool you already use.
Is anything saved?
No. Nothing is uploaded or stored, and there is no account.