Midjourney AI Image Generator

Midjourney V8.2 is the current default image model for bold visual exploration, personalized aesthetics, and reference-led editing. Use Ottermind to organize the brief, prompt variants, and source images, then generate through Midjourney's official website or Discord.

Skincare UGC

Mango campaign

Perfume short

Skincare story

Midjourney V8.2 Image Generation at a Glance

Released on July 24, 2026, V8.2 is Midjourney's current default image model. It focuses on aesthetics, image quality, and more responsive Personalization, while the new Edit Model supports written edits, inpainting, outpainting, and up to four reference images. At a square aspect ratio, V8.2 outputs 1024 x 1024 SD images or native 2048 x 2048 HD images. Midjourney does not offer a general public API, so Ottermind Studio is the connected workspace for planning prompts and references, comparing exported results, and deciding the next revision rather than an unofficial Midjourney endpoint.

Key Features of Midjourney AI

Text and Image Prompts: Build a new image from a written scene, one visual reference, or multiple images combined with descriptive text.
Style Reference: Carry colors, medium, texture, and lighting from a reference image or style code without asking the model to copy its subject.
Personalization and Moodboards: Use rated preferences or a curated image collection to guide the visual character of repeated generations.
V8.2 Edit Model: Edit an existing image, combine up to four references, change selected regions, or expand the canvas with written direction.
SD and HD Output: Generate a 1024-pixel square in SD or a native 2048-pixel square in HD, with supported ratios varying by mode.

What Midjourney AI Does Best

Midjourney V8.2 is strongest when visual taste, atmosphere, composition, and rapid exploration matter as much as literal instruction following. Build a precise brief in Ottermind Studio, generate several candidates through an official Midjourney surface, and bring the selected images back for comparison and follow-on work.

Campaign concepts and art direction
Explore distinct lighting, styling, composition, and mood directions before a team commits to one visual system for a launch or social campaign.
Editorial product imagery
Develop premium-looking still-life concepts, material studies, and background treatments, then use an AI Product Image Generator workflow when the brief needs structured product inputs and iterations.
Posters, covers, and visual worlds
Generate striking key art, narrative environments, and stylized subjects. For layout-led work, carry the chosen visual direction into an AI Poster Generator and check every word separately.

How Midjourney V8.2 Compares with Earlier Midjourney Models

Best for

Midjourney V8.2: Current default generation, updated aesthetics, stronger Personalization, and the new reference-led Edit Model.
Midjourney V8.1: Fast V8-series image generation and native HD output with the earlier V8.1 aesthetic behavior.
Midjourney V7: Legacy V7 workflows that depend on Omni Reference, Draft Mode, or an established V7 look.

Release status

Midjourney V8.2: Released July 24, 2026 and currently the default Midjourney image model.
Midjourney V8.1: Released April 14, 2026; selectable, but no longer the default.
Midjourney V7: Released April 3, 2025; selectable as an older version.

Reference workflow

Midjourney V8.2: Image Prompts, Style References, Moodboards, Personalization, and Edit Model with up to four attached references.
Midjourney V8.1: Image Prompts, Style References, Moodboards, Personalization, and the same current Edit Model.
Midjourney V7: Image Prompts, Style References, Moodboards, Personalization, and one Omni Reference for a character or object.

Editing

Midjourney V8.2: Edit Model handles written changes, multi-image composition, inpainting, outpainting, and retexturing.
Midjourney V8.1: Compatible with the current Edit Model and its reference-led editing workflow.
Midjourney V7: Pan, Zoom Out, and region edits use legacy paths; Omni Reference is not compatible with several editing actions.

Square output

Midjourney V8.2: 1024 x 1024 SD or native 2048 x 2048 HD before any later upscale.
Midjourney V8.1: 1024 x 1024 SD or native 2048 x 2048 HD before any later upscale.
Midjourney V7: No native V8-style HD generation; V7 uses its documented upscale workflow.

Aspect ratio

Midjourney V8.2: Up to 14:1 in SD and 4:1 in HD, with extreme ratios treated as experimental.
Midjourney V8.1: Up to 14:1 in SD and 4:1 in HD, with extreme ratios treated as experimental.
Midjourney V7: Up to 14:1 in the current version compatibility table.

How to choose

Midjourney V8.2: Start with V8.2 for new work, especially when Personalization or multi-reference editing matters.
Midjourney V8.1: Compare V8.1 when you prefer its look or need a stable baseline from an existing V8.1 project.
Midjourney V7: Return to V7 only when its aesthetic or Omni Reference behavior is part of an established workflow.

Spec	Midjourney V8.2	Midjourney V8.1	Midjourney V7
Best for	Current default generation, updated aesthetics, stronger Personalization, and the new reference-led Edit Model.	Fast V8-series image generation and native HD output with the earlier V8.1 aesthetic behavior.	Legacy V7 workflows that depend on Omni Reference, Draft Mode, or an established V7 look.
Release status	Released July 24, 2026 and currently the default Midjourney image model.	Released April 14, 2026; selectable, but no longer the default.	Released April 3, 2025; selectable as an older version.
Reference workflow	Image Prompts, Style References, Moodboards, Personalization, and Edit Model with up to four attached references.	Image Prompts, Style References, Moodboards, Personalization, and the same current Edit Model.	Image Prompts, Style References, Moodboards, Personalization, and one Omni Reference for a character or object.
Editing	Edit Model handles written changes, multi-image composition, inpainting, outpainting, and retexturing.	Compatible with the current Edit Model and its reference-led editing workflow.	Pan, Zoom Out, and region edits use legacy paths; Omni Reference is not compatible with several editing actions.
Square output	1024 x 1024 SD or native 2048 x 2048 HD before any later upscale.	1024 x 1024 SD or native 2048 x 2048 HD before any later upscale.	No native V8-style HD generation; V7 uses its documented upscale workflow.
Aspect ratio	Up to 14:1 in SD and 4:1 in HD, with extreme ratios treated as experimental.	Up to 14:1 in SD and 4:1 in HD, with extreme ratios treated as experimental.	Up to 14:1 in the current version compatibility table.
How to choose	Start with V8.2 for new work, especially when Personalization or multi-reference editing matters.	Compare V8.1 when you prefer its look or need a stable baseline from an existing V8.1 project.	Return to V7 only when its aesthetic or Omni Reference behavior is part of an established workflow.

Where Midjourney AI Stands Out

Strengths creators can use

Aesthetic range and visual impact: V8.2 is built for bold, polished image exploration across photographic, illustrative, cinematic, and design-led directions.
Layered style control: Style References, style codes, Personalization profiles, and Moodboards give creators several ways to carry a visual language across prompts.
Reference-led editing: The V8.2 Edit Model can combine up to four references and supports instructed changes, selected-region edits, and canvas expansion.

Where creators still need to refine

Exact instruction following: A visually compelling result may still omit a requested object, rearrange a composition, or favor atmosphere over a literal commercial brief.
Typography and fixed layouts: Check every character, line break, logo shape, and reserved text area before using an output in a poster, package, or ad.
Edit-region stability: Early creator reports on V8.2 describe changes outside the intended edit area, so compare each revision with its source instead of assuming untouched regions stayed identical.
No general public API: Use Midjourney's official website or Discord. Unauthorized automation and third-party access are prohibited by Midjourney's current guidelines.

Creator Feedback on Midjourney AI

Recent creator discussions do not point to one universal verdict. V8.2 is praised for speed, atmosphere, and improved handling of longer prompts, while other creators still prefer an earlier version for a particular look or report inconsistency in exact edits. Treat these as recurring workflow observations rather than guaranteed model behavior.

Visual polish remains the main attraction

Creators repeatedly highlight cinematic lighting, editorial finish, rich atmosphere, and strong first-pass concepts as reasons to use Midjourney.

Prompt accuracy still needs active review

Longer prompts can hold together better in successful V8.2 examples, but strict object counts, text, and functional layout requirements still benefit from several controlled variations.

Editing is useful but newly changing

The multi-reference Edit Model expands what can be done in one workflow, yet creators are still testing how reliably it preserves details across repeated local edits.

How to Create with Midjourney AI and Ottermind

Build the brief

Collect the deliverable, subject, composition, style references, aspect ratio, and review criteria in Ottermind Studio.

Generate officially

Take the approved prompt and references to midjourney.com or the official Discord bot, choose V8.2 and the required SD or HD mode, then generate comparable candidates.

Compare and refine

Bring selected exports back to the same Ottermind thread, note adherence and visual issues, and continue with an AI Image Editor or a revised Midjourney prompt.

Midjourney AI FAQs

What is the latest Midjourney image model?

Midjourney V8.2 is the current default image model as of September 3, 2026. It was released on July 24, 2026 and focuses on aesthetics, image quality, Personalization, and the new Edit Model.

Can Midjourney V8.2 use reference images?

Yes. Use Image Prompts to influence content and composition, Style References for visual treatment, Moodboards or Personalization for broader aesthetic direction, and the Edit Model for instructed editing or composition with up to four reference images.

What resolution does Midjourney V8.2 generate?

At the default square ratio, V8.2 generates 1024 x 1024 pixels in SD or 2048 x 2048 pixels in HD. Other aspect ratios change the pixel dimensions, and HD supports a narrower maximum ratio than SD.

Does Midjourney have an official public API?

No general public API is documented. Midjourney's current community guidelines say that, except for rare explicitly granted cases, it does not provide an API and prohibits unauthorized automation or third-party apps.

Can I use Midjourney without Discord?

Yes. Every current subscription plan supports image creation on the Midjourney website as well as in Discord. A Discord account is only required when you choose the Discord workflow.

Does Midjourney V8.2 replace Omni Reference?

For V8.x workflows, Midjourney directs creators to the Edit Model, which replaces Omni Reference and Character Reference and accepts up to four reference images. Omni Reference remains a V7 feature.

Can I generate Midjourney images directly in Ottermind?

Not through an unofficial integration. Use Ottermind to organize the brief, prompt versions, source images, exports, and review notes, then run the actual generation on Midjourney's official website or Discord surface.

Should I use Midjourney V8.2, V8.1, or V7?

Start new projects with V8.2. Compare V8.1 when an existing project depends on its output style, and use V7 when its particular aesthetic or Omni Reference behavior is necessary. Keep the prompt and seed conditions comparable when testing versions.

Plan a Better Midjourney AI Workflow

Keep the creative brief, reference images, prompt versions, exported candidates, and review decisions in one Ottermind workspace, then use Midjourney through its official creation surfaces.