Don't get left behind in the AI revolution!

Stay ahead of the curve with our newsletter packed with expert AI tips and tricks, the latest AI news and trends, and exclusive discounts and offers.

    Get 100+ Free AI Tools to Boost Your Productivity

    Want to work smarter, not harder?
    Let the AI work for you.

      Image Generator

      Midjourney

      1.69K
      Please log in or register to do it.

      Midjourney is a subscription AI image generator known for its polished, art-directed look. As of September 2026 its default model is V8.2, with an instruction-based Edit model, image-to-video, Draft mode and deep style controls. Plans run from $10 to $120 per month, with no free trial on the website.

      Last updated: September 2026

      What is Midjourney?

      Midjourney is made by Midjourney, Inc., a San Francisco company founded by David Holz, co-founder of Leap Motion. It launched its Discord server in March 2022 and entered open beta on 12 July 2022. A full web app followed in August 2024, and midjourney.com is now the main place to create, while the Discord bot remains available.

      The current default model is V8.2, released on 24 July 2026. Midjourney says it makes images “more creative, bold, sophisticated, edgy and fresh”, dramatically reduces random low-quality results, and improves Personalization. It follows V8.1 (released 14 April 2026, default from 10 June), which Midjourney calls its fastest model, with standard jobs rendering about 4–5 times faster than earlier versions. V7 was the default from June 2025 to June 2026. For anime and illustration there is a separate Niji 7 model (January 2026), though V8.1 and V8.2 have no Niji version.

      When Meta agreed in August 2025 to license Midjourney’s image and video technology, Holz said the company remains independent with no investors, and TechCrunch noted that Midjourney has never taken outside funding.

      Key features of Midjourney

      V8.2 image model and HD output

      You type a prompt and Midjourney returns a grid of four images. Since V8.1 you can generate HD images at 2K resolution without a separate upscale step. According to the Version documentation, an HD job costs 1.3 minutes of GPU time versus 0.8 minutes for standard definition. V8.2 supports the familiar parameters, including Chaos, Raw, Stylize, Tile, Weird and Seed, plus Relax and Fast modes. Turbo mode is not supported in V8.1 or V8.2.

      Edit model and the web Editor

      Midjourney released its V8.2 Edit model for public testing on 27 August 2026. It lets you:

      • Edit an image with plain instructions (for example, “make it night”)
      • Generate with up to four reference images, replacing Omni Reference and Character Reference in V8.x
      • Inpaint and outpaint selected areas in the Editor
      • Combine edits with Personalization and moodboards

      The Editor is web-only and includes an Erase brush, Smart Select, layers, Pan and Zoom Out, and Retexture, and it works on your own uploaded photos as well as Midjourney images. A 24 September 2026 update said targeted edits now change only the selected pixels. On Discord, the Edit model is available through the –edit parameter.

      Video generation

      Midjourney’s V1 Video Model (18 June 2025) turns a still image into video. Click Animate on any image, or upload your own as a start frame, and choose automatic or manual motion and low or high motion. Each clip is 5 seconds and can be extended four times, by about 4 seconds each, up to 21 seconds. You can also set an end frame or make loops. SD video is 480p; HD is 720p. At launch Midjourney priced a video job at roughly eight times an image job.

      Draft and Conversational modes

      Draft mode is for fast, cheap exploration. In V8.1 and V8.2 a Draft prompt costs 0.4 minutes of GPU time and returns 24 images at 512×512 pixels, but it only works on midjourney.com, not Discord. Conversational mode lets you describe changes in plain language (“run a variation of image 4”), and voice input is available when Draft mode is on.

      Style references, Style Creator and moodboards

      • Style Reference (–sref): copies the colors, textures, medium and lighting of an image, not its objects. Style weight (–sw) runs from 0 to 1000, and –sref random draws from Midjourney’s internal style library.
      • Style Creator: a web-only tool that builds a custom –sref code as you pick from grids of samples. Most styles settle after 5–10 rounds, and the preview images use your GPU time.
      • Moodboards: collections of your own uploaded, linked or gallery images that become a reusable style, applied with –p and a moodboard ID.

      Personalization

      By selecting the images you like on the Personalize page, you build a profile of your taste, then apply it with –p. Midjourney’s docs say a V7 Global Personalization Profile works with V8.2, and the V8.2 announcement says profiles now draw on a larger pool of images.

      Stealth mode and privacy

      By default, Midjourney images are public and can be remixed by other users. Stealth mode, available only on Pro and Mega, hides your creations on the website. Images made in public Discord channels stay visible even in Stealth mode.

      Benefits of using Midjourney

      • Strong default aesthetics: short prompts still produce finished, stylized images, and V8.2 targets fewer weak results.
      • Style consistency: sref codes, the Style Creator, moodboards and Personalization make it easier to keep a visual identity across a campaign.
      • One subscription for images and video: every plan can animate images, and Pro and Mega can generate SD video in Relax mode.
      • Predictable monthly cost: flat plans, with unlimited Relax-mode images from Standard up.

      Midjourney use cases by role

      • Designers and art directors: moodboards, concept art and style exploration in Draft mode before a final HD render.
      • Marketers and social media teams: campaign visuals kept on-brand with a saved sref code or moodboard, plus short animated clips for social posts.
      • Founders and small businesses: product mockups, pitch-deck imagery and website hero images without a photo shoot.
      • Illustrators and game artists: character and environment concepts, anime styles with Niji 7, and seamless textures with –tile.

      How much does Midjourney cost?

      The table below shows Midjourney’s plans as of September 2026, taken from its plans documentation. Prices are in USD. Annual billing saves 20%, and every plan can buy extra Fast GPU time at $4 per hour.

      PlanMonthlyAnnualFast GPU timeRelax modeStealth mode
      Basic$10$96/year ($8/month)3.3 hours/monthNoNo
      Standard$30$288/year ($24/month)15 hours/monthUnlimited imagesNo
      Pro$60$576/year ($48/month)30 hours/monthUnlimited images and SD videoYes
      Mega$120$1,152/year ($96/month)60 hours/monthUnlimited images and SD videoYes

      Basic runs 3 concurrent Fast jobs; Pro and Mega run up to 12. HD video is available on Standard, Pro and Mega in Fast mode only. There is no free trial on the website or Discord; the only trial is a limited one in the niji·journey app for iOS and Android.

      Commercial use terms

      Midjourney’s Terms of Service (effective 27 May 2026) say you own the assets you create “to the fullest extent possible under applicable law”, and that ownership continues if you downgrade or cancel. A company, or an employee of a company, with more than $1,000,000 in annual revenue must subscribe to Pro or Mega to own its assets. You also grant Midjourney a broad license to your content, and upscaled images of other users remain owned by their original creators.

      Midjourney pros and cons

      Pros

      • Distinctive, high-quality aesthetics with little prompt engineering
      • Deep style tools: sref codes, Style Creator, moodboards and Personalization
      • Instruction-based Edit model with up to four reference images
      • Image-to-video included in every plan
      • Flat pricing with unlimited Relax-mode images from $30 per month

      Cons

      • No free trial on the website or Discord
      • Images are public by default; privacy needs the $60 Pro plan or higher
      • Frequent changes: V8.x dropped Omni Reference and Character Reference, and the Edit model is still in testing
      • Video is image-to-video only and tops out at 720p
      • Ongoing copyright lawsuits from major studios and artists

      Legal matters to know

      On 11 June 2025, Disney and NBCUniversal sued Midjourney in the US District Court for the Central District of California in Los Angeles, alleging it trained on and reproduces characters such as Darth Vader, Elsa and the Minions. Warner Bros. Discovery filed a similar suit in September 2025 covering characters including Superman, Batman and Scooby-Doo. Midjourney argues that fair use in generative AI is unsettled law; The Art Newspaper reported that on 29 June 2026 it asked the court to make the studios disclose their own internal AI use, after an earlier request was denied in mid-June. Midjourney is also a defendant in Andersen v. Stability AI, an artists’ lawsuit filed in January 2023.

      How to use Midjourney

      1. Go to midjourney.com, sign in and subscribe to a plan (Basic is the cheapest; choose Pro if you need Stealth mode or your company earns over $1 million a year).
      2. Optional: build your taste profile on the Personalize page, or create a moodboard from reference images.
      3. On the Create page, type a prompt in the Imagine bar. Turn on Draft mode to explore ideas cheaply.
      4. Pick the best result and render it in HD, or add –sref, –p or other parameters to steer the style.
      5. Use the Editor or attach up to four reference images to make instruction-based edits, inpaint or extend the canvas.
      6. Click Animate to turn a finished image into a 5-second video, then extend it if needed.
      7. Manage your creations on the Organize page and download the files you need.

      Midjourney alternatives

      • ChatGPT: pick it for conversational image creation and editing inside a general assistant; OpenAI launched ChatGPT Images 2.0 in April 2026 with better text rendering and followed with Images 2.5 in September.
      • Google Gemini: pick it for Nano Banana 2, Google’s fast image model with precise text rendering, consistency for up to five characters and output up to 4K; it is free in the Gemini app, with some features reserved for paid tiers.
      • Adobe Firefly: pick it if commercial safety matters most; Adobe says its Firefly models are trained on licensed and public-domain content, and it also offers partner models.
      • Ideogram: pick it for posters, logos and graphics with legible text; Ideogram 4.0 is open-weight and has a free web tier.
      • FLUX: pick it if you want open weights or API access to build image generation into your own product with Black Forest Labs’ models.

      Frequently asked questions

      Is Midjourney free?

      No. According to Midjourney’s documentation, there is no free trial on midjourney.com or Discord, and you need a paid plan to generate images. The only trial is a limited one in the niji·journey mobile app for iOS and Android. The cheapest plan is Basic at $10 per month, or $8 per month billed annually.

      What is the latest Midjourney version?

      V8.2 is the default model. It became the default on 24 July 2026, replacing V8.1, which had been the default since 10 June 2026. V8.2 focuses on aesthetics, fewer low-quality results and better Personalization. V7, released in April 2025, was the default from June 2025 until June 2026.

      Can Midjourney make videos?

      Yes. Midjourney’s V1 Video Model, released on 18 June 2025, animates an image into 5-second clips that you can extend up to four times, to a maximum of 21 seconds. SD video is 480p on every plan. HD 720p video needs a Standard, Pro or Mega plan and runs in Fast mode only.

      Can I use Midjourney images commercially?

      Yes. Midjourney’s Terms of Service (effective 27 May 2026) say you own the assets you create to the fullest extent possible under applicable law, and ownership continues if you later cancel. Companies, and employees of companies, with more than $1,000,000 in annual revenue must use the Pro or Mega plan to own their assets.

      Are my Midjourney images private?

      Not by default. Midjourney’s terms state that content is publicly viewable and remixable unless you use Stealth mode, which is only available on the Pro and Mega plans. Even with Stealth mode on, images made in public Discord channels remain visible, so use the website, Discord direct messages or a private server.

      What happened to Omni Reference and Character Reference?

      In V8.1 and V8.2 they are replaced by the Edit model, which Midjourney released for public testing on 27 August 2026. It accepts up to four reference images and supports instruction-based edits, inpainting and outpainting. Omni Reference still exists, but only for V7, where it costs twice the GPU time of a normal image.

      Should I use Midjourney on Discord or the website?

      The website is now the main interface. Midjourney’s getting started guide is written for midjourney.com, and several tools, including the Editor, the Style Creator and V8 Draft mode, are web-only. Discord still works for prompting, the Edit model (via the –edit parameter) and community support.

      Verdict

      Midjourney remains a top choice in September 2026 when image quality and a consistent visual style matter more than anything else. V8.2, the new Edit model and built-in video make it more complete than a year ago. Standard at $30 per month suits most regular users; businesses earning over $1 million, or anyone who needs private images, should budget for Pro. If you need a free tool, reliable text in images or strict commercial safety, look at the alternatives above.

      Sources

      Craiyon AI
      DALL·E 3
      Ad Area

      Reactions

      0
      0
      0
      0
      0
      0
      Already reacted for this post.

      Reactions