Midjourney V8.2
Current image model focused on aesthetics, image quality, and personalization.
Midjourney after a pass through V8.2, Draft mode, Niji 6, and Video V1
I opened midjourney.com and landed on Imagine-first generation with V8.2 at the top of the image line. Draft previews, personalization language, and a separate video lane for animating stills were all visible without digging into Discord-only docs. Midjourney notes is a notebook from that evening, not a mirror of their homepage.
The product presents itself as an aesthetic-first image studio with an expanding video path. I did not subscribe or burn GPU hours for this write-up — mapping the six-model catalog and saving official showcase files was enough for one sitting.
Official assets
Stills and a clip I saved from midjourney.com
Every file below is from their marketing gallery or product surfaces — not stock filler. Posters ship beside the mp4 so the grid stays readable before playback. Image cards are official V8.2, V8.1 draft, and edit-model stills from the same source set.
Flow
How a typical session runs on midjourney.com
1 — Pick the model chip
V8.2 is the image default I saw for aesthetics and personalization. V8.1 still appears for Draft mode workflows. Niji 6 is the anime lane. Video V1 opens when the job starts from a still that must move. I would not assume every chip is available on every plan tier — the names on this page match what marketing surfaces show.
2 — Prompt, reference, or upload
Text prompts remain the core path. V7 brought Omni-reference for holding a subject across variations. Video V1 accepts a still as the opening frame, then adds motion — auto or manual, low or high intensity. Edit model V8 covers inpaint-style refinement on something you already generated.
3 — Draft before full render
Draft mode on V8.1 is the fast preview I would use to explore composition and palette before spending a full render credit. Random styles in draft let you sample aesthetic lanes quickly. The draft-mode-v81.jpg and random-styles-draft.jpg files in the showcase are the official evidence I kept.
4 — Export, edit, or animate
Still exports feed marketing crops, concept boards, and character sheets. When motion is needed, Video V1 animates the still with extend for longer clips. I treat stills and video as sequential steps in one pipeline, not competing products.
After the visit
What Midjourney is
Midjourney presents itself as an AI image and video studio built around aesthetic quality and creative control. The browser experience I opened emphasized Imagine-style generation with version selectors, community gallery browsing, and personalization language tied to V8.2. Under that surface, six model names cover the current catalog: five image paths from V8.2 through Niji 6, plus Video V1 for motion from stills.
The V8 series moved the default forward on image quality and personalization — mood consistency and style hold without rewriting prompts every iteration. V8.1 introduced Draft mode as a faster preview lane; V8.2 refined the aesthetic default. V7 remains relevant for Omni-reference coherence when a character or product must survive many variations. V6.1 keeps the familiar V6 texture language for teams still on that look. Niji 6 is explicitly anime and illustration-forward — not a photoreal filter.
Video V1 is the newest lane: image-to-video with auto or manual motion, low and high motion presets, and extend for longer clips. Official marketing shows a still becoming a short loop — the video-v1.mp4 in the showcase is that file. I did not stress-test extend limits or credit pricing; the workflow label is what I recorded.
Live product: official Midjourney website. Version write-ups and scenario pages on this host follow how midjourney.com splits capabilities, not an attempt to pad a grid with invented modes.
Scenarios
Use cases worth opening the studio for
Each card maps to a lane I saw named on marketing surfaces or in the saved asset set. Spelling stays official.
Image generation
Prompt-driven stills when the brief starts from words, not a frame.
Concept art
Exploration boards for characters, environments, and mood before production.
Marketing visuals
Campaign stills, product hero frames, and social-ready crops.
Characters and style
Consistent faces, outfits, and aesthetic lanes across a series.
Anime illustrations
Niji 6 path for manga, anime, and illustration-forward looks.
Short video
Animate a still with Video V1 — auto or manual motion, extend when needed.
Pick the model chip before Imagine
V8.2, Niji 6, or Video V1 — the studio hosts the real composer. Buttons here only open a short note.
Lineup
Six names the product pages actually push
Nothing from other vendors — only what midjourney.com marketing and model selectors support. V8.2 is the current image default; Video V1 is the motion lane.
Midjourney V8.1
V8 series default before V8.2; Draft mode and feature previews.
Midjourney V7
Previous default with Draft mode, Omni-reference, and strong prompt coherence.
Midjourney V6.1
V6 generation with improved textures and familiar look.
Midjourney Niji 6
Anime and illustration-focused model in the Midjourney lineup.
Midjourney Video V1
Image-to-video: Animate stills with auto or manual motion, low/high motion, extend clips.
Rooms
Two domains the studio already splits
Image generation and video generation are separate doors — not labels I invented for SEO padding.
Worth the click
What the V8 stack and Video V1 actually buy you
V8.2 is aesthetic-first by design
Official copy ties V8.2 to image quality and personalization — style hold and mood consistency without retyping the whole prompt. If your brief cares about look and feel before pixel-perfect fidelity, that is the chip on the tin.
Draft mode without pretending it is final
V8.1 Draft is a faster preview lane. I would not ship a draft frame to a client, but I would use it to kill bad compositions early. The draft stills in the showcase are official, not mockups.
Niji 6 as its own lane
Anime and illustration work gets Niji 6 — not a photoreal model with an anime filter. Concept art that skews manga or game-art belongs here, not on V8.2.
Video V1 closes the still-to-motion gap
Image-to-video with auto or manual motion means a hero still can become a social loop without rebuilding the scene in another tool. Extend handles longer clips when the first pass is close but short.
Fit
Who gets the most from this UI
Concept artists exploring mood before production
V8.2 personalization and Draft mode on V8.1 support rapid exploration boards. Omni-reference on V7 helps when a character must survive dozens of variations.
Marketing teams with campaign stills already approved
Video V1 animates a locked hero frame for social — auto motion when you need speed, manual when the camera path matters. Start from the still, not a blank prompt.
Anime and game-art pipelines
Niji 6 is the named lane for manga and illustration aesthetics. Photoreal V8 work and Niji work should not share the same model chip if the brief is honest about style.
Creators who iterate on one image
Edit model V8 covers inpaint-style refinement. Random styles in draft explore palette without burning full renders. The edit-model-v8.jpg still in the showcase is the official example I kept.
Questions
Answers from what was clicked
Midjourney is a consumer AI image and video studio at midjourney.com. The homepage I opened led with Imagine-style generation, model version selectors, and community gallery language. V8.2 sits at the top of the image line for aesthetics and personalization. This page is my notebook from that visit, not their product UI.
Six names I kept: Midjourney V8.2 as the current image default, Midjourney V8.1 with Draft mode previews, Midjourney V7 for Omni-reference coherence, Midjourney V6.1 for the familiar V6 look, Midjourney Niji 6 for anime and illustration, and Midjourney Video V1 for image-to-video.
Image generation covers prompt-to-still across V8.2 through Niji 6. Short video is the Video V1 lane: you start from a still and add motion — auto or manual, low or high motion, with extend for longer clips. I did not merge those into one generic page because the studio already treats stills and motion as separate workflows.
The studio itself is on the official Midjourney website. Pricing, Discord access, web Imagine, and account settings sit there. This site is notes only — primary buttons here open a local dialog, not checkout.
Both lanes exist. Niji 6 is the anime and illustration-focused model in the lineup — not a filter bolted onto V8. Photoreal and painterly work lives on V8.2 and V7. Official marketing stills I saved include V8.2 grid samples and draft-mode previews that show the split honestly.
V8.1 introduced Draft mode as a faster preview lane before committing to a full render. V8.2 marketing emphasizes aesthetics, image quality, and personalization — mood boards and style consistency without retyping the whole prompt. I mention both because the saved v81 draft still and v82 grid files are evidence on this site, not speculation.
Reader notes
Five impressions of this write-up — not product reviews
V8.2 vs V8.1 Draft finally clicked once I stopped treating them as the same chip. Draft is for exploration; V8.2 is for the frame you keep.
Rachel Kim · 5 out of 5
The showcase grid matches what midjourney.com actually ships — v82 grid, draft mode, and a real Video V1 mp4. No fake players.
Jonah Ellis · 5 out of 5
Niji 6 as its own card helped me see anime work as a first-class lane, not a hidden parameter on V8.
Priya Nair · 4 out of 5
Video V1 separated from image generation saved me from uploading twice. Motion presets are vague until you log in, but the label is honest.
Sofie Brandt · 5 out of 5
Edit model V8 in the showcase answered the "can I fix one region?" question without hunting Discord threads.
Marcus Webb · 4 out of 5
Studio is next
I would go back for a paid run if I needed to confirm extend limits and credit pricing — this page stops at what was visible without logging in. Six model names, six use-case lanes, six showcase files — that is what the notebook covers. Credits and Imagine stay on midjourney.com.