Y
Prompt Workshop
Article · Midjourney parameters

Midjourney --chaos Parameter Explained

--chaos (0–100) controls how different the four candidate images Midjourney generates are from each other. It is the parameter for the exploration phase.

What --chaos does

Every Midjourney generation returns 4 candidate images. With --chaos 0, all four are near-identical variations on the same composition. With --chaos 100, they can be wildly different interpretations of your prompt.

5 reference values

--c valueSpread across 4 imagesBest for
0 (default)Near-identicalStable, reproducible work
10–25Composition variationSame direction, different framing
50Distinct interpretationsDirection exploration
75–100Wildly differentBrainstorming, unexpected angles

Standard exploration workflow

  • Round 1: --c 50–75 with a loose prompt to find a direction
  • Pick the strongest candidate, note its prompt and seed
  • Round 2: --c 0 with the same seed to refine
  • Final: --c 0 with chosen --s to lock the result
Pro tip: "High --chaos during exploration, low --chaos during convergence" is the standard professional workflow.

Quick comparison

--c 0Stable
cyberpunk samurai standing in neon rain, cinematic --ar 16:9 --v 6 --c 0

Four near-identical variations.

--c 50Exploration
cyberpunk samurai standing in neon rain, cinematic --ar 16:9 --v 6 --c 50

Four distinctly framed variations.

--c 100Brainstorm
cyberpunk samurai standing in neon rain, cinematic --ar 16:9 --v 6 --c 100

Four candidates in completely different visual languages.

Frequently asked questions

What is the difference between --chaos and --weird?

--chaos controls inter-candidate variance; --weird controls within-image weirdness. They serve different purposes.

Does --chaos 100 hurt image quality?

No. --chaos only affects spread between candidates, not quality.

How does --chaos pair with --stylize?

Exploration: --c 50 --s 200–300. Convergence: --c 0 --s tuned to the scenario.

Try this parameter live in the editor

The local editor includes Midjourney / SDXL / Flux templates with a visible, editable parameter panel.

Open the editor →
Yan · AI Prompt Workshop editorial team|Last updated on 2026-06-12。This site does not call any cloud model. Every prompt and parameter in this article was tested and refined locally by the editorial team.