Project MonetRequest demo
Home/Blog/Midjourney V8.2 Edit Model: Image Editing, Inpainting & How It Works

Creator Tools · Project Monet Briefing

Midjourney V8.2 Edit Model: What It Can Do and How to Use It

Midjourney's first V8.2-native edit model adds instruction-driven editing, multi-image references, inpainting and outpainting in a broad public test.

Published 2026-08-29 · Updated 2026-08-29 · By Project Monet Editorial Team

Project Monet editorial graphic showing four image references feeding into a before-and-after generative edit with inpaint and outpaint boundaries

01

What the V8.2 Edit Model is

Midjourney began broad testing of its first V8.2 image-edit model on August 27, 2026. Midjourney explicitly describes the release as a test with fast-follow updates, not a finished stable rollout.

The model can edit existing images from natural-language instructions and can also generate with multiple image references instead of treating editing as only a legacy side workflow.

02

Instruction editing and image references

Midjourney says the model supports instruction-based editing and up to four image references at once. The company positions the multi-reference workflow as a replacement for omni-reference in this editing path.

It also supports personalization, moodboards and style references. Midjourney cautions that moodboards and srefs may benefit from extra prompt direction in the current test version.

03

Inpainting and outpainting

Inpainting targets a selected region for change while attempting to preserve the surrounding composition. Outpainting expands the canvas and asks the model to generate plausible content beyond the original frame.

For creators and marketers, that can support campaign variants, product-image restyling, social aspect-ratio changes, wardrobe or environment changes and extra negative space for copy. Generative edits still need human review for logos, typography, faces and product geometry.

04

How to access the test

Midjourney lists several access paths: attach images in the prompt bar, use Edit from the lightbox, upload through the Edit tab, or use --edit with an image URL in Discord.

The company says both its main and alpha web interfaces are being updated around the test, with the alpha interface expected to change rapidly while new interaction patterns are tried.

05

How this differs from the older Editor

Midjourney already had an Editor interface for repainting, reframing, pan, zoom and vary-region workflows. The August 27 announcement is specifically about the first V8.2-native image-edit model.

Do not assume every older Editor control has already migrated to the new model or behaves identically. Midjourney's own documentation should be rechecked while the test evolves.

06

Current limitations

The biggest limitation is maturity: Midjourney says there will be edge cases and rapid interface/model updates. A successful edit may also change nearby style, composition or identity beyond the intended instruction.

Sources

Primary and supporting sources

Facts were rechecked against the linked sources immediately before publication. Pricing, product availability and rollout status can change.

Project Monet

Useful signals. Clear decisions. Better digital work.

Project Monet turns relevant shifts in AI, creator tools and the web into practical context—and builds focused websites for businesses ready to grow.

Request a free homepage concept