Image models

Midjourney tests a Thinking Mode that reruns images to fix prompt misses

Midjourney is testing Thinking Mode on its alpha site: rerun a V8.2 image and it looks for where the result missed your prompt, then generates again.

Midjourney is testing a Thinking Mode for image generation on its alpha website, alpha.midjourney.com. Its October 7, 2026 changelog calls it an early version and asks for feedback.

What to know

  • You can rerun jobs made with V8.2, standard and edit, using “Rerun (Thinking)” under the lightbox.
  • Midjourney says it looks at your image, finds where it missed your prompt (wrong objects, layout, anatomy or text) and generates again.
  • In an earlier post, Midjourney said its tests show it helps with prompt accuracy, typography and coherence, and that it might offer it more broadly in future.
  • It is a test on the alpha site, so it may change or not reach everyone.

Also in the same changelog

  • Shared folders: invite people as Collaborators, who can generate into the folder, or as Viewers, who can browse it. Midjourney warns that sharing does not change image privacy, and non-stealth images can still appear on Explore and profiles.
  • Folders can be pinned to the sidebar, and large moodboard uploads now show a progress bar.

See our Midjourney review and Midjourney parameters explained.

Source: Midjourney updates. We summarise in our own words; check the source for the full details.

Related

More AI news

stay in the loop

New guides, prompts and AI model news by email