AI image generation tool that creates stunning artwork, illustrations, and photorealistic images from text prompts.
Midjourney is one of the most popular AI image generation platforms, known for producing exceptionally high-quality and aesthetically striking visuals. Describe what you want in a text prompt and Midjourney generates images in seconds — from fantasy illustrations and architectural renders to photorealistic portraits and abstract art. With a web-based editor, you can upscale, create variations, remix, and fine-tune results.
Used by designers, artists, marketers, and creative professionals worldwide for concept art, mood boards, marketing assets, and creative exploration.
Pricing model: Paid
Common questions about Midjourney
Midjourney’s alpha site now has shared folders. Invite people by username as Collaborators who can generate into the folder, or Viewers who can browse its images. Link sharing can limit access to invitees, allow anyone with the link to view, or let any Midjourney member contribute.
Shared folders are an early version. Sharing doesn’t change image privacy: non-stealth images can still appear on Explore and profiles, even inside a private folder.…
Midjourney is testing Thinking Mode on its Alpha website. Open an existing job’s lightbox and click “Rerun (Thinking)” in the lower right corner beneath it.
Midjourney says its tests show improvements in prompt accuracy, typography, and coherence. It’s asking testers to share examples of where the mode helps or falls short in ideas-and-features. Broader availability remains a possibility while testing continues.
Original source
Midjourney's alpha site now groups all the style references in a prompt into a single style pill, with thumbnails and codes you can see on hover.
Sending a previewed style keeps your prompt-bar settings, and prompt parameters stay saved per folder within a tab's session. Edits also keep the base image's aspect ratio, so a stray --ar won't stretch the result.
This round is mostly fixes and cleanup. Collaborative tools were planned for the following week.
Original source
Midjourney's alpha site now lets you preview your prompt across styles as you browse. Turn on "Live previews" under the Styles tabs and scroll. The previews use fast models and your prompt text alone; clicking a style sends your prompt with that style applied.
You can also save default parameters through Settings → Advanced → Your defaults, or choose "Set current pills as default" in the prompt bar.
The Create feed now uses a full-width masonry layout, with prompts and buttons on hover.…
Midjourney is testing live style previews in its alpha interface. Turn on “Live previews” in the Styles sidebar to see your latest prompt across liked and featured styles, then click a thumbnail to generate with that style.
The edit model's inpainting and outpainting now change only selected pixels. Midjourney says you can make repeated edits without degrading image quality. It also says --tile results on V8.1 and V8.2 now blend without visible seams.
Read the original announcement →
Midjourney's alpha site now supports Korean and includes another round of editor and mobile fixes.
In the v8.2 editor, draft batches show source images in a grid with hover previews, and a working Enhance option replaces the old Upscale button. Add to Folder also works when you don't have any folders yet.
On mobile and tablet, pages use the full width, menus stay on screen, and uploads have a 3-dot menu for deleting them.…
Midjourney's alpha has a lightbox editor powered by the v8.2 edit model. Open an image, edit it with plain-text instructions, attach up to 4 reference images and see your session's edits together.
The editor launched the week before, and Midjourney says this update fixes bugs and glitches. It's also testing an early Change Style feature in the lightbox.
This is an alpha release, and Midjourney warns that some things may be broken.
Read the original announcement →
Midjourney is collecting ideas for what to work on next. You can submit suggestions at midjourney.com/ideas, with a formal voting session planned for the next week or two.
Read the original announcement →
Midjourney says its updated V8.2 edit model improves image quality. It's asking anyone who had edit issues in the previous 24 hours to try again and send feedback.
Read the original announcement →
Midjourney has updated its V8.2 edit model for what it says is better image quality. If you hit problems in the previous 24 hours, Midjourney is asking you to retry and send feedback.
Other products to explore and follow.
The people behind Midjourney
Founder
Top-ranked AI image platform for photorealism, vector generation, custom styles, mockups, and more — built for designers and teams.
The lab behind FLUX — frontier open and commercial models for image, and now video, audio, and action prediction in one multi-modal architecture.
AI design tool that creates fully-editable, on-brand slides, social posts, PDFs, and diagrams on a real canvas you control.
ByteDance's AI-native cloud platform — tap the AI powering TikTok, including Seedance video and Seedream image models, LLMs, and enterprise APIs.
AI-powered design tool by Replit — describe what you want and get professional, brand-consistent designs in minutes.
AI-powered UI prototype generator — create, iterate, and test designs that match your design system, from a prompt.
