V8.1 · Native Upscale
A storm you can wear
A translucent perfume vessel becomes a tiny weather system, with electric clouds, glass detail, and a creature-like silhouette.
Model
Image Prompts(0/20)
Add reference images
PNG, JPEG, WebP · max 10MB each
Specialized References
Guide style, character identity, or a specific object.
Style References(0/20)
Add reference images
PNG, JPEG, WebP · max 10MB each
Character References(0/20)
Add reference images
PNG, JPEG, WebP · max 10MB each
Object Reference(0/1)
Add reference images
PNG, JPEG, WebP · max 10MB each
0 / 1024
Privacy ModePremium
Credits required: 10
Create a four-image Midjourney set
Start with a prompt, then add image, style, character, or object references when you need more control.
V8.1DraftRawStyle Reference4 Results
No history yet
Midjourney AI Image Generator with V8.1 and References
Create, compare, and refine four Midjourney images in one browser workspace, from the first prompt to references, variations, editing, and upscaling.
Flux Context brings Midjourney into the same media library and credit workflow as your other image, editing, and video tools. New creators get a focused starting point, while experienced Midjourney users can open familiar controls when they need them.
V8.1 + Niji 7Draft from 5 credits4 images per requestStyle, Character & Object ReferencesVary + Native Upscale
Why use Midjourney in Flux Context?
This hosted workflow focuses on moving creative work between models and tools. It does not replace Midjourney's own community or account features.
One browser workspace
Create with Midjourney, edit with other image models, upscale a final selection, or continue into image-to-video without manually moving files between services.
Visible cost per action
See the exact credit requirement before generation, Image Edit, Vary, or native Midjourney Upscale is submitted.
Shared media and creation history
Reuse uploaded assets and find running or completed tasks in My Creations instead of managing temporary image links.
Cross-model continuation
Use a selected Midjourney result as the input for Flux Context's editor, upscaler, or other creative models.
Privacy controls
Eligible accounts can enable Flux Context Privacy Mode from the same generator before submitting a task.
Recoverable async tasks
Temporary provider or network delays remain processing, while My Creations and tab resume checks help recover the final result.
Midjourney speed, mode, and credit guide
The generator calculates the final price from the selected mode, speed, and HD option before you submit.
| Workflow | Availability | Credits | Best for |
|---|---|---|---|
| Draft | V7 Create | 5 | Low-cost exploration |
| Fast | Create | 10 | Balanced iteration |
| V8.1 HD Fast | Text-to-image | 15 | 2K final candidates |
| Turbo | Create | 20 | Short feedback loops |
| Image Edit | Fast / Turbo | 15 / 30 | Full-image transformation |
| Vary | Selected result | 15 | Subtle or strong alternatives |
| Native Upscale | Selected result | 15 | Midjourney follow-up processing |
A Midjourney workflow that grows with your prompt
Start with the essentials, then bring in the reference and aesthetic controls that match your creative direction.
Midjourney V8.1 and Niji
Choose the latest general model or a familiar previous and anime-focused version without rewriting version flags in your prompt.
- •V8.1 is the default
- •V7 Draft for lower-cost exploration
- •Niji 7 and Niji 6 for anime and illustration
Version compatibility is enforced before credits are deducted.
Four Reference Workflows
Separate visual composition, style, character identity, and object guidance instead of treating every upload as the same kind of reference.
- •Image Prompts for composition and visual content
- •Style References for a reusable visual direction
- •Character and Object References with weight controls
Use up to 20 references in one request.
Simple by Default, Detailed When Needed
Start with version, speed, ratio, and prompt. Open Advanced Settings only when you want direct control over the Midjourney aesthetic.
- •Stylization, Variety, and Weirdness
- •Raw and Negative Prompt controls
- •Reference-specific weight sliders
Unsupported combinations are disabled instead of silently submitted.
Select, Refine, and Continue
Work from a four-image result grid and keep the selected image moving through the workflow without copying image URLs by hand.
- •Vary Subtle and Vary Strong
- •Native Midjourney Upscale
- •Reuse any result as the next Image Prompt
Every follow-up action shows its credit cost before submission.
How the reference types work
Assigning each image a purpose gives you more predictable control than placing every upload into one undifferentiated list.
Image Prompt
Guide the subject, composition, color, or visual content of the new image.
Style Reference
Carry a visual treatment, medium, lighting language, or design direction into the result.
Character Reference
Guide recurring character identity while your prompt changes the scene or treatment.
Object Reference
Use one reference to guide a product, prop, vehicle, creature, or other recognizable object.
Real Midjourney results
Strange enough to stop the scroll
Every image below was generated with the Midjourney versions and follow-up actions available on this page. The collection deliberately favors surprising subjects over generic sample art.
V8.1 · Fast
The greenhouse astronaut
Porcelain, moth wings, and botanical growth turn a familiar space portrait into an uncanny editorial character study.
V8.1 · HD Fast
A flower that watches back
HD output keeps the mechanical filaments, translucent petals, and specimen-table textures readable at a larger scale.
Niji 7 · Fast
Koi above an inverted city
An anime-focused pass turns impossible architecture, airborne fish, and saturated light into a graphic cinematic frame.
Style + Character + Object References
References, recombined
Separate reference roles guide the visual language, character direction, and featured object into one emerald-and-gold synthesis.
Selected result → Vary Strong
Keep the idea, change the surprise
Select one image from the four-image grid, then ask for a stronger variation. The second result keeps the storm-vessel idea while pushing its creature form, lighting, and composition in a new direction.
Original: V8.1 FastVary Strong: 15 creditsNative Upscale: 15 credits
Want to build from one of these directions?
Start at the generator, then use references and follow-up actions to make the result your own.
From first prompt to a finished asset
1
Choose a workflow
Start with Create or upload one base image for full-image editing.
2
Add direction
Write the prompt, choose V8.1 or another version, then add references only when they improve control.
3
Compare four results
Review the full set and select the composition worth continuing.
4
Refine or continue
Run Vary, use native Midjourney Upscale, reuse the result as a reference, or open another Flux Context tool.
Keep working after Midjourney
A selected image can move directly into Flux Context's regular image upscaler or natural-language editor. You can also browse other models when the next step needs different strengths.
Midjourney AI Image Generator FAQ
What Midjourney versions are available?
The generator supports Midjourney V8.1, V7, V6.1, V6, Niji 7, and Niji 6. V8.1 is the default. Draft is available with V7, while optional 2K HD output is available with V8.1 Fast text-to-image.
How many credits does Midjourney use?
Draft costs 5 credits, Fast costs 10 credits, V8.1 HD Fast costs 15 credits, and Turbo costs 20 credits for text or reference-guided creation. Image Edit costs 15 credits in Fast or 30 credits in Turbo. Vary and native Midjourney Upscale cost 15 credits each.
What reference image types can I use?
You can use ordinary Image Prompts plus dedicated Style, Character, and Object References. A request can contain up to 20 references in total, with one Object Reference. Weight controls appear in Advanced Settings when the matching reference type is present.
Can I edit an existing image with Midjourney?
Yes. Choose Edit Image, upload exactly one base image, and describe the final result. This first version performs full-image editing. Region masks and canvas editing are not currently exposed.
What is the difference between Midjourney Upscale and Upscale Image?
Midjourney Upscale is a native follow-up action tied to a selected result and costs 15 credits. Upscale Image opens Flux Context's separate lower-cost image upscaler, which is useful when you only need resolution enhancement.
Is Flux Context the official Midjourney website?
No. Flux Context is an independent, third-party creative platform that provides a hosted Midjourney workflow alongside other image, editing, upscaling, video, and music models. Midjourney's own website remains the place for its community, Moodboards, personalization, and other native account features.
Where can I find a generation after leaving the page?
Running and completed tasks are saved to My Creations. The generator also checks the task again when you return to a visible or online tab, so a temporary network interruption does not require another paid submission.
Create your next Midjourney image set
Start with V8.1 and a simple prompt, or bring references and advanced controls into the same four-image workflow.