# Midjourney Online - Version, Niji, and Style Controls > Canonical: https://stable-diffusion-web.com/midjourney ## Midjourney - configurable text and reference image form The current form supports text or one reference image and exposes nine ratios, quality, version, Niji, HD, stylize, chaos, weirdness, output count, and paid-model access. ## Midjourney version and style examples The gallery illustrates concept-art, anime, fantasy, landscape, surreal, and fashion directions; it is not a benchmark or an aesthetic, identity, or quality guarantee. ### Cinematic Concept Art ### Anime Niji Style ### Fantasy Character ### Painterly Landscape ### Surreal Art ### Fashion Editorial ## Why Midjourney The live form owns the current prompt, reference, ratio, quality, version, Niji, HD, stylize, chaos, weirdness, and output-count controls. Review every returned image. ### Text-led and one-image modes Text mode starts from a required prompt. Image mode requires one JPEG, PNG or WebP reference up to 10 MB plus a prompt. ### Four version choices The version selector offers 8.1, 7, 6.1, and 6. Compare returned results rather than transferring release or quality claims into the labels. ### Niji and HD controls Niji offers Off, 0, 5, 6, and 7; HD is a separate switch. These fields are controls, not guarantees of style or detail. ### Stylize, chaos, and weirdness Stylize spans 0-1000, chaos 0-100, and weirdness 0-3000. Change one variable at a time to understand its effect. ### Nine aspect ratios Choose 1:1, 9:16, 16:9, 4:3, 3:4, 2:3, 3:2, 9:21, or 21:9. ### Paid access with configured estimates The catalogue requires paid-model access; metadata starts at 3 credits and estimates about 180 seconds. Confirm the live debit before submitting. ## Who uses Midjourney's version, Niji and style controls? Concept artists, illustrators, designers, marketers, and editorial teams can use these candidate workflows to evaluate version, Niji, HD, style, ratio, and output-count behavior; every result still needs review. ### Concept-art comparisons Hold the brief constant while comparing version or style controls, then inspect composition, anatomy, detail, and artifacts. ### Anime and illustration studies Test Niji choices as selectable form options and reject identity, hand, text, clothing, or background drift. ### Editorial visual drafts Generate candidate layouts, then verify likeness, rights, disclosure, text, and brand details before publication. ### Style-control experiments Compare stylize, chaos, or weirdness with one field changed at a time and record which output fits the brief. ### Fantasy and environment concepts Explore world-building directions while checking perspective, subject count, depth, geometry, and visual coherence. ### Reference-led compositions Use one source image to condition composition or style, while treating it as guidance rather than a subject or identity lock. ## What a run costs and returns | Measure | Detail | | --- | --- | | 1 model | Available in the form today, each priced in the same credits. | | 20 credits | Per run, depending on the model and length chosen. | | Every paid plan | Includes a commercial licence for what you make. | | JPEG, PNG, and WebP | Accepted for upload, up to 10 MB per file. | | 2,000 characters of prompt | The longest prompt every model on this page accepts. | ## Midjourney version, Niji and style FAQ Answers about text and one-image input, versions, Niji, HD, quality, stylize, chaos, weirdness, ratios, output count, paid access, credits, time, rights, and review limits. ### What is Midjourney on this page? Midjourney is a selectable paid-access image workflow in this app with text-led and one-image modes. The live form defines its version, Niji, HD, quality, style, ratio, and output-count controls. ### Which input modes are available? Text mode uses a required prompt. Image mode requires one JPEG, PNG or WebP reference up to 10 MB plus a prompt. ### Which versions and Niji choices can I select? Versions are 8.1, 7, 6.1, and 6. Niji choices are Off, 0, 5, 6, and 7. Compare outputs rather than assuming a label fits every task. ### Which quality and HD controls are exposed? Quality offers 0.25×, 0.5×, 1×, and 2×, while HD is a separate switch. Inspect the returned file and updated debit. ### What ranges do stylize, chaos, and weirdness use? Stylize spans 0-1000 in steps of 10, chaos 0-100 in steps of 5, and weirdness 0-3000 in steps of 10. ### Which aspect ratios can I choose? The form offers 1:1, 9:16, 16:9, 4:3, 3:4, 2:3, 3:2, 9:21, and 21:9. ### How much does a run cost, and what access is required? The catalogue marks the model as paid access and metadata starts at 3 credits with an estimated 180 seconds. Confirm the live estimate before submitting. ### Can paid-plan output be used commercially? Paid plans include this app's commercial license. You remain responsible for prompt and reference rights, applicable terms, and final use. ### What should I review before using an image? Review prompt adherence, identity, anatomy, text, geometry, crop, artifacts, reference drift, and delivered files. Generated images are candidates rather than verified final assets. ### How many credits does Midjourney use? Midjourney draws 20 credits per run. ### What limits should I check before using Midjourney? Generated output from Midjourney can contain visual, audio, or factual errors. Review each result before publishing, and upload only material you are authorized to use. Where the form accepts a reference image, supported files are JPEG, PNG, WebP up to 10 MB each. ### What happens to the files I upload to Midjourney? An upload is stored against your own account, not pooled with other people's. Deleting the account removes it: the control is in Settings under Security, the request is confirmed by email, and the erasure removes the stored objects it owns. How long anything else is kept, and which processors handle it, is stated in the privacy policy. ### Can I use what Midjourney makes commercially? Every paid plan includes a commercial licence for what Midjourney makes; the free tier is for personal use. ### How does Midjourney compare, and what alternatives fit a different workflow? Midjourney: Use Midjourney with text or one JPEG, PNG or WebP reference, nine ratios, quality, version, Niji, HD, stylize, chaos, weirdness, output count, and paid access. Related options are Stable Diffusion Models Hub: Compare the Stable Diffusion models and the other image and video models here: the inputs each takes, what a run costs, and the plan it needs, Midjourney Video: Use Midjourney Video with one required source image, 480p or 720p output, six aspect ratios, low or high motion and paid access, Kling AI: kling ai: run Kuaishou's Kling 3.0 online. Turn text or an image into 3-15s cinematic clips with native audio-visual sync, intelligent storyboard, Nano Banana: Use Nano Banana with a prompt or up to four JPEG, PNG, or WebP references, ten ratios, Auto framing in image mode, and one to four outputs, Hailuo 02: Use Hailuo 02 with text-led or image-led modes, 02 or 2.3 image versions, Standard or Pro, and 6- or 10-second choices where compatible, and Nano Banana 2: Use Nano Banana 2 with text or up to four JPEG, PNG, or WebP references, 1K, 2K or 4K, ten ratios, Auto framing in image mode, and one to four outputs. Choose by the required input, delivered output, controls, limits, current availability, and credit cost; the related-tools section links each canonical page. ## Test a current Midjourney workflow Open the generator, confirm paid-model access, choose text or image mode, set the displayed controls, review the estimate, and inspect every result. ## Related tools ### [Stable Diffusion Models Hub](https://stable-diffusion-web.com/models) ### [Midjourney Video](https://stable-diffusion-web.com/midjourney-video) ### [Kling AI](https://stable-diffusion-web.com/kling) ### [Nano Banana](https://stable-diffusion-web.com/nano-banana) ### [Hailuo 02](https://stable-diffusion-web.com/hailuo) ### [Nano Banana 2](https://stable-diffusion-web.com/nano-banana-2) _Last updated July 14, 2026_