Midjourney¶

Development line: project:midjourney · thread midjourney-model-and-feature-releases
Last event: 2025-04-04 · 6 dated since 2022-07-12 · Researched: 2026-09-04 · confidence: medium

What it is¶

Midjourney is a subscription image-and-video generator for art direction, concept work, and rapid visual iteration.

• Generates stills from text, image prompts, style references, personalization, and Edit Model references.
• Supports variations, inpainting, outpainting, canvas expansion, and turning a still into a 5-second video.

Development line¶

• 2022-07-12 — Midjourney beta access documented. The archive recorded a beta-access item with a shortened link, but the link leaves eligibility, scope, and behavior unverified.
• 2022-11-07 — Midjourney V4 introduced. The archive linked the official site and external coverage of Midjourney V4, without establishing exact features or access conditions.
• 2023-01-16 — Midjourney Blend command documented. The archive recorded a Blend command item, but kept no source URL to verify behavior or rollout.
• 2023-03-09 — Midjourney opened V5 staging. Output became more photographic and matched prompts more closely, though longer prompts were sometimes needed for the intended aesthetic.
• 2023-03-16 — Midjourney V5 documented. The archive recorded a V5 milestone, but kept no URL to confirm timing, features, or availability.
• 2025-04-04 — Midjourney released V7 Alpha. The archive linked the official V7 Alpha update page and official Imagine page, but the links do not establish full feature sets or access rules.

What changed¶

On 2022-07-12, the beta-access note lacked accessible corroboration; official history places V3 as the default model from July 2022. On 2022-07-26, no URL or product change survived. On 2022-11-07, V4 introduced a new codebase and AI architecture, improving coherence, detail handling, image prompting, and multi-prompts after its November 2022 release. On 2023-01-16, a Blend event was recorded without an independent launch date; /blend remains available in Discord for combining 2–5 images. On 2023-03-09, V5 staging aligned with the March 2023 V5 release, offering more photographic output and closer prompt matching, with longer prompts needed for some aesthetics. On 2023-03-16, no surviving URL separated this mention from the March V5 rollout. On 2023-05-03, a community report alleged new V5.1 prompt blocking; official history dates V5.1 to May 4, 2023, introducing stronger default aesthetics, coherence, natural-language interpretation, and --tile, leaving the blocking policy unverified. On 2023-08-22, no surviving URL supports a product change. On 2025-04-04, V7 Alpha opened with default personalization, Draft Mode, and improved prompt/image coherence, while some editing functions temporarily fell back to V6. On 2025-06-17 (found today), V7 became the default model. On 2026-04-14 and 2026-06-10 (found today), V8.1 launched and then became the default, adding faster generations and optional 2K HD images. On 2026-07-24 (found today), V8.2 became the default, and its Edit Model replaced V7 Omni Reference, Character Reference, and Retexture in the V8 workflow.

How to use this¶

As of 2025-04-04, treat model versions and alpha releases as distinct workflow targets: confirm availability through the official update and Imagine pages before adopting a version in production.

1. Sign in or create an account, pick a subscription plan, and open the Create page.
2. Enter a concise visual prompt in the Imagine bar to generate an initial set of images as described in Creating on Web.
3. Use V8.2 as the default, or pin a model with --v # via Version when repeatable behavior matters.
4. Supply up to four reference images or edit an image using written text with the Edit Model.
5. Pick a solid result, then generate Subtle or Strong Variations instead of starting over.
6. For motion, turn a still into the initial frame of a 5-second Video; select motion and plan settings deliberately because video consumes more GPU time.

Best practices¶

• Write short, concrete visual descriptions covering only key subject, medium, setting, lighting, color, mood, and composition details, following Prompt Basics.
• Treat Image Prompts as influence rather than exact copies: describe every required element in text, and use the Editor for fine edits.
• Use a Style Reference to set appearance, and write the subject as fresh content rather than an instruction to modify the reference.
• Use --seed to compare prompt adjustments against an identical V8 starting point per Seeds; do not use seeds as character or style locks.
• Build a Personalization profile or Moodboard per Personalization for recurring project styles; curated image selections yield a tighter match.
• Upload only images you own the rights to, and keep external edits compliant with service rules and applicable law under Editor.

Superseded by this¶

• 2022 beta and V3 access rules are obsolete for daily use: current work starts with a subscription and the web Create page or Discord bot.
• Guidance naming V4, V5, or V5.1 as the default model is superseded by V8.2; pin --v only when legacy output is required.
• 2023 Blend instructions do not apply to the web interface: pass multiple image prompts without text on web; /blend remains Discord-only.
• Guidance for V7 Omni Reference, V6 Character Reference, and Retexture is superseded by the V8 Edit Model.
• V7 as the default model is superseded by V8.2; V7 remains a selectable legacy model.

Still unknown¶

• The claims behind the 2022-07-12 beta-access note and unlinked entries on 2022-07-26, 2023-03-16, and 2023-08-22 cannot be reconstructed from available records.
• The short link, V5 staging page, Discord message, Reddit post, and DTF page were unreachable during research and do not serve as factual support.
• Current documentation dates the V7 release to 2025-04-03, while the V7 Alpha announcement is dated 2025-04-04; the reason for this one-day gap remains unresolved.
• Current subscription prices and usage allowances were not gathered, so no specific tier is recommended.

Sources¶

Agent brief¶

• Subject: project:midjourney, thread midjourney-model-and-feature-releases, 6 dated events 2022-07-12 → 2025-04-04.
• Practical note: As of 2025-04-04, practitioners should treat Midjourney model versions and alpha releases as distinct workflow targets: confirm current availability through the official update and Imagine pages before adopting a version in a production workflow.
• Confidence: medium. Dated supersedes above are the authority for what is obsolete.