Midjourney is an independent AI research lab whose flagship product turns text prompts into highly aesthetic images, widely regarded as among the most visually striking of any image generator. It runs primarily through its web app at midjourney.com (Discord is an alternative) with subscription plans from $10/month, and it has expanded into video generation alongside images. All plans include commercial usage rights, but there is no free tier and private 'Stealth mode' generations require the Pro or Mega plan.
Key features
Prompt-based text-to-image generation via the midjourney.com web app
AI video generation (SD resolution on Basic, SD and HD on higher plans)
Relax mode for unlimited slower generations on Standard, Pro and Mega plans
Stealth mode for private generations on Pro and Mega plans
Repeat and permutation jobs for batch variations (up to 40 jobs on Pro/Mega)
General commercial usage rights included on every plan
Fast GPU time allocations from 3.3 to 60 hours per month with $4/hr top-ups
Watch: Midjourney tutorial
FULL 2026 Midjourney AI Tutorial (for Beginners) — Wade McMaster - Creator Impact
Video by Wade McMaster - Creator Impact on YouTube — plays here without leaving this page.
Midjourney models, modes & features
No models or features match your search.
Models
V8.2
Model
Current default Midjourney image model. An update focused on aesthetics, image quality and Personalization; it introduced the new Edit Model, which replaces Omni Reference, Character Reference and the Retexture tool. V8.2 images are described as more creative, bold, sophisticated and edgy, and Personalization profiles draw on a much larger pool of images.
Access
Any paid subscription; selected with --v 8.2 or as the default in settings
Released
2026-07
Focus on aesthetics, image quality and PersonalizationNew Edit Model for editing and reference workflowsNative HD images (2K, 2048px) without upscalingMax aspect ratio 14:1 (4:1 for HD)Style Reference, Image Prompts, Remix, Pan and Zoom Out supported
Limits: Became the default version on July 24, 2026 · Omni Reference and Character Reference are not supported — replaced by the Edit Model · Turbo Mode is not supported · Quality parameter is not supported
Released April 14, 2026. Midjourney's fastest model generation, with better prompt adherence and native HD (2K) image generation. Served as the default version from June 10 to July 23, 2026.
Access
Any paid subscription; selected with --v 8.1
Released
2026-04
Fastest Midjourney model at its release — standard jobs render about 4–5x faster than earlier versionsNative HD 2K images without upscaling (--sd / --hd; HD costs 1.3 GPU minutes vs 0.8 for SD)Better prompt reading and small-detail retention; Raw mode removes default styling for more prompt adherenceNew Edit Model supported
Limits: Was the default version from June 10 to July 23, 2026; superseded as default by V8.2 · Turbo Mode is not supported · May require unlocking a Global V7/V8 Personalization profile · Inpainting/outpainting on HD images downscales results to SD (upscalers can restore HD)
Early-access alpha of the V8 family, launched March 17, 2026 on the Midjourney Alpha website. Retired on July 24, 2026 when V8.2 became the default.
Access
Was available only on the Midjourney Alpha website
Released
2026-03
Early preview of the V8 model family
Limits: Early-access alpha launched March 17, 2026 · No longer available for use after July 24, 2026; images already generated remain in the website gallery
Released April 3, 2025 and the default model from June 17, 2025 to June 9, 2026. Handles text and image prompts with high precision, with richer textures and more coherent details — especially in bodies, hands and objects. Introduced Draft Mode and Omni Reference.
Access
Any paid subscription; selected with --v 7
Released
2025-04
Precise handling of text and image promptsRicher textures and more coherent details, especially bodies, hands and objectsIntroduced Draft Mode for fast, cheaper iterationIntroduced Omni Reference (--oref) for character/object referenceConversational Mode supported
Limits: Was the default version from June 17, 2025 to June 9, 2026 · Multi-Prompting is not supported · Uses the V6.1 upscalers and V6.1-based Editor/Pan/Zoom Out
Released July 30, 2024 and the default version until June 16, 2025. Produces more coherent images with more precise details and textures, and generates images approximately 25% faster than V6.
Access
Any paid subscription; selected with --v 6.1
Released
2024-07
More coherent images with more precise details and textures than V6Still powers the upscalers and editing tools used with V6 and V7
Limits: Was the default version until June 16, 2025 · Generates images approximately 25% faster than V6 (per official docs)
Released December 20, 2023 and the default version from February 14 to July 30, 2024. Brought enhanced prompt accuracy for longer inputs, improved coherence and knowledge, and advanced image prompting and remixing capabilities.
Access
Any paid subscription; selected with --v 6
Released
2023-12
Enhanced prompt accuracy for longer inputsImproved coherence and knowledgeAdvanced image prompting and remixingSubtle and Creative upscalers
Limits: Was the default version from February 14 to July 30, 2024 · Older architecture — fewer features than V7/V8 (no Draft Mode, Omni Reference, Edit Model or HD images)
Launched January 9, 2026. The current Niji model — a special series developed with Spellbrush for Eastern and anime aesthetics and illustrative styles, with a major boost in coherency and prompt following.
Access
Any paid subscription; selected with --niji 7 (uses the V7-era feature set; not available on V8)
Released
2026-01
Eastern and anime aesthetics and illustrative stylesMajor coherency boost — clearer eyes, reflections and small background detailsCloser prompt following for specific designs and repeatable charactersCleaner, flatter look highlighting improved line work
Limits: More literal than earlier Niji versions — broad or 'vibey' prompts may behave differently · Separate Niji Journey website and Discord server
Midjourney's video model: turns a single image into a 5-second video, extendable up to 21 seconds. Per the official Video documentation, all plan tiers can generate videos in Fast Mode, Standard/Pro/Mega can also generate HD (720p) videos in Fast Mode, and Pro/Mega can generate SD videos in Relax Mode.
Access
Any paid subscription (web app or Discord); HD requires Standard, Pro or Mega; Relax video requires Pro or Mega
Image-to-video: animate a Midjourney image or an uploaded image used as the starting frame5-second videos, extendable in +4-second steps up to a maximum of 21 secondsAuto, Manual and Loop animation options; Low Motion and High Motion settings (--motion low / --motion high)End frames via --loop and --end parametersBatch sizes of 1, 2 or 4 videos (--bs); a batch of 4 costs 8 GPU minutes in SD or 26 GPU minutes in HDDownload as MP4 (social or raw) or GIF
Limits: A starting image is required — it animates images rather than generating video from text alone · SD output is 480p; HD output is 720p and only in Fast Mode on Standard, Pro and Mega plans · Relax Mode video is SD only and limited to Pro and Mega plans · Style Reference, Image Prompt and Edit Model Reference are not compatible with video generations
Editing and reference system introduced with V8.2 (also on V8.1). On V8 versions it replaces Omni Reference, Character Reference and the Retexture tool, and is what the Editor, Pan and Zoom Out tools use.
Access
V8.1 and V8.2
Replaces Omni Reference, Character Reference and the Retexture tool on V8Powers Pan, Zoom Out and the Editor on V8 versionsUse images and plain text instructions to create or modify images
Limits: Not available on V6 or V7 (per the official feature compatibility chart)
Midjourney's image editor for modifying generated or uploaded images. Listed as supported across V6, V7 and V8 in the official feature compatibility chart.
Access
All versions
Edit and repaint regions of an imageOn V6 and V7 it runs using V6.1; on V8 it runs using the Edit Model
Limits: Working behavior depends on the selected model version
Upscaling tools that enlarge finished images. The official compatibility chart lists Subtle & Creative upscalers for V6, the V6.1 upscalers for V7, and SD-image upscaling for V8.1/V8.2.
Access
All versions
Upscale generated images to higher resolutionSubtle and Creative upscalers on V6; V7 uses the V6.1 upscalers; V8 upscales SD imagesCan return edited HD images to HD resolution
Feature/parameter family for guiding generations with a reference image's style. Style Reference, Style Reference Codes and Style Weight are supported across V6, V7 and V8 per the official compatibility chart. A parameter feature, not a model.
Access
V6, V7 and V8
Match the look and feel of a reference imageStyle Reference codes for reusable stylesStyle Weight parameter (--sw) controls influence
Omni Reference (--oref) / Character Reference (--cref)
Feature
Reference parameters for reusing a subject across images: Character Reference (--cref) on V6, generalized to Omni Reference (--oref) in V7 (characters, objects and more in one parameter), and replaced by the Edit Model in V8.2. Parameter features, not models.
Access
V7 (--oref); V6 used --cref; V8 uses the Edit Model instead
Keep a character, object or vehicle consistent across generations
Limits: Omni Reference is V7-only in the official compatibility chart · Character Reference (--cref) is not supported on V7/V8; on V8.2 it is replaced by the Edit Model
Personalization tunes Midjourney's output toward your taste using profiles built from images you rate and pick; Moodboards package a set of reference images into a reusable style. Both are supported across V6, V7 and V8.
Access
All versions
Personalization profiles learn your aesthetic from rated imagesMoodboards curate a set of images into a reusable style profileV8.2 Personalization 'understands' aesthetic tastes better, with a larger image pool for building profiles
Limits: V8.1 use may require unlocking your Global V7/V8 Personalization profile
Outpainting tools that expand an image's canvas. Supported on all current versions per the official compatibility chart.
Access
All versions
Extend an image beyond its original borders in any directionZoom Out expands the scene at the same center image
Limits: On HD images, Pan and Zoom Out downscale the result to SD (upscalers can restore HD) · On V8 these tools run through the Edit Model; on V7 through V6.1
Privacy feature that keeps your generations out of Midjourney's public community surfaces. Available only on the Pro and Mega plans per the official plan comparison.
Access
Pro and Mega plans only
Keeps your images and videos private instead of publicly viewable/remixable by default
Limits: Creations made in shared or open spaces such as a public Discord chatroom remain viewable there regardless of Stealth Mode
Generation speed modes that determine how jobs are queued and billed. Fast uses your plan's Fast GPU hours; Relax offers unlimited but slower queued generation on higher plans; Turbo trades more GPU time for speed on older versions.
Access
All plans (Relax availability depends on plan and job type)
Fast Mode: priority generation consuming Fast GPU timeRelax Mode: unlimited queued generation on eligible plans (images on Standard+; SD video on Pro/Mega)Turbo Mode: faster generation at higher GPU cost on V6/V7
Limits: Turbo Mode is not supported on V8.1/V8.2 · Basic plan has no Relax GPU time
Midjourney does not offer an official public API. The official documentation covers only the midjourney.com web app and the Discord bot, both behind a paid subscription, and lists no developer endpoints or API keys. The official Terms of Service (version effective May 27, 2026) further prohibit programmatic access: 'You may not use automated tools to access, interact with, or generate Assets through the Services.' Services advertised elsewhere as a 'Midjourney API' are unofficial third-party wrappers, not a Midjourney product.
API pricing: No API pricing exists: there is no official API, no API keys and no usage-based developer billing. Access is sold only as consumer subscriptions (Basic/Standard/Pro/Mega) used through the midjourney.com web app and the Discord bot.
Integrations & platforms
Midjourney web app (midjourney.com Create page)Discord (Midjourney Bot)Niji Journey website and Discord server (for Niji models)
What should you use it for?
Highest-quality, on-trend still images
Use the current default V8.2 for the best aesthetics, image quality and Personalization, with native HD 2K output.
Anime and manga-style illustration
Use Niji 7 (--niji 7) — built with Spellbrush for anime aesthetics, with clearer fine details and closer prompt following for repeatable characters.
Keeping the same character or product consistent across images
Use V7 with Omni Reference (--oref), or V8.2 with the Edit Model, which replaced Character Reference.
Fast creative ideation
Use V7's Draft Mode to explore compositions quickly and cheaply before committing Fast GPU time to full renders.
Turning artwork into short video clips
Animate a finished image with the Video model (5s clips, extendable to 21s; HD on Standard/Pro/Mega, Relax SD video on Pro/Mega).
Private or client work that must not appear publicly
Subscribe to Pro or Mega for Stealth Mode — the only plans that keep images and videos private.
High-volume image generation on a budget
Use Standard or higher for unlimited Relax Mode image generation once Fast hours run out; Basic has no Relax time.
Pros
Official positioning centers on 'the most beautiful AI models in the world' — consistently top-rated for artistic quality
Fast generation with clear GPU-time allowances and an unlimited slower Relax mode on most plans
Commercial usage rights included even on the cheapest plan
Cons
No free plan or trial — the entry point is a $10/month subscription
Private generations (Stealth mode) require the $60/month Pro plan or higher; lower plans are public by default
No official public API, so programmatic access requires unofficial wrappers
Model, plan and pricing information on this page was last verified on 22 October 2026 against official sources. Prices and line-ups change — always confirm on the official website before buying.
The official docs' Video article documents the video model in detail but does not state a release date and does not itself use the 'V1' name in the text reviewed; 'Video V1' is the name used in Midjourney's launch communications, so 'released' is null here.
V8.2's Version article entry states only that it 'released as the default version on July 24, 2026'; no separate earlier release date is given, so 2026-07 is recorded.
Versions older than V6 (V1–V5.2, Niji 5/6) are covered in the docs' separate Legacy Features article, which was not opened in this pass; only V6 onward and Niji 7 are listed to avoid citing unopened material.
No official API documentation exists to cite for the API section; the absence of an API is established from the official Terms of Service prohibition on automated tools and from the official docs covering only the web app and Discord bot.
GPU-minute costs vary by job type and settings; only the per-job costs stated on the official Version and Video pages (e.g., HD batch costs, video batch costs) are reflected.
The niji・journey mobile app is referenced in third-party coverage but was not verified on an official page opened in this pass, so integrations list only the Niji Journey website/Discord named in the official Version article.
How to choose the right Image & Design tool
For image and design tools, judge by output quality on your kind of project — product shots, social posts, illustrations or brand assets — not by gallery examples. Control matters: can you set sizes and aspect ratios, keep a consistent style or character, edit one part of an image without regenerating everything, and export in the formats you need (PNG, SVG, print resolution)? Check the licence carefully before commercial use: who owns the output, and are stock elements included? Also compare pricing honestly — credit-based plans can get expensive fast if you iterate a lot, so a free trial with your real workload tells you more than any feature list.
Midjourney is an independent AI research lab whose flagship product turns text prompts into highly aesthetic images, widely regarded as among the most visually striking of any image generator. It runs primarily through its web app at midjourney.com (Discord is an alternative) with subscription plans from $10/month, and it has expanded into video generation alongside images. All plans include commercial usage rights, but there is no free tier and private 'Stealth mode' generations require the Pro or Mega plan.
What are the main features of Midjourney?
Midjourney offers Prompt-based text-to-image generation via the midjourney.com web app, AI video generation (SD resolution on Basic, SD and HD on higher plans), Relax mode for unlimited slower generations on Standard, Pro and Mega plans, Stealth mode for private generations on Pro and Mega plans, Repeat and permutation jobs for batch variations (up to 40 jobs on Pro/Mega), General commercial usage rights included on every plan and Fast GPU time allocations from 3.3 to 60 hours per month with $4/hr top-ups.
How much does Midjourney cost?
Midjourney is a paid product. Prices and plans change over time, so check its official website for the current pricing before you buy.
What are the pros and cons of Midjourney?
Strengths of Midjourney include Official positioning centers on 'the most beautiful AI models in the world' — consistently top-rated for artistic quality; Fast generation with clear GPU-time allowances and an unlimited slower Relax mode on most plans; Commercial usage rights included even on the cheapest plan. On the downside, No free plan or trial — the entry point is a $10/month subscription; Private generations (Stealth mode) require the $60/month Pro plan or higher; lower plans are public by default; No official public API, so programmatic access requires unofficial wrappers.
Offline AI image search and photo manager for Windows — describe a meme, screenshot, receipt or family photo (or search the OCR text inside it) and find it across all folders without a filename, with all AI running locally on your PC
Best for: Windows users with large photo, screenshot and document libraries who want private, offline content search
Pricing:Free
Limitation: Windows only — no macOS or Linux version on the official site
AI visual generator with a layer editor — generate a deck or visual, then edit it instead of regenerating it (brush-repaint areas, split a flat image into editable layers, pull text off an image to retype it), with brand assets applied throughout and open-source agent skills so Claude Code or Cursor can call it from the terminal
Best for: Designers and agent-workflow builders who want editable, publish-ready AI visuals instead of one-shot generations
Limitation: No pricing or plan details published on the site yet — the maker only says 'try it free'
AI CAD generator that turns a prompt or reference image into an editable assembly and CAD that can be inspected, edited and refined, with all export formats including STEP
Best for: Designers and makers who want editable CAD from a prompt or reference image
Pricing:Freemium
Limitation: Free plan capped at 200 credits a month — the Beta credit gift is one-time only
Welcome to LLM Tools Hub! Hi! I'm Hubi 👋 I help you find the right AI tool from LLM Tools Hub. Tell me what you need — for example: “free AI tools for image generation”
