Midjourney

Vendor: Midjourney

Image

Latest Midjourney v8.2 image generation with follow-up actions

Typical latency
30-90s
Reference inputs
Up to 4
toolId
midjourney

Input modes

text-to-imageimage-to-image

Variants

Imagine

Prompt and image-reference generation on Midjourney v8.2.

text-to-imageimage-to-image

Edit

Text-guided edits for one or more reference images.

image-to-image

Blend

Blend two to four uploaded images into a new composition.

image-to-image

Describe

Generate prompt ideas from one uploaded image.

image-to-image

Upscale

Upscale one image from a completed four-image Midjourney grid.

text-to-image

Variation

Create a variation from a selected image in a completed grid.

text-to-image

Vary Strong

Create a stronger variation from an upscaled image.

text-to-image

Vary Subtle

Create a subtle variation from an upscaled image.

text-to-image

Reroll

Rerun a completed Midjourney task with the same prompt.

text-to-image

Zoom Out

Zoom out from an upscaled Midjourney image.

text-to-image

Pan

Extend an upscaled image left, right, up, or down.

text-to-image

Inpaint

Start an inpaint job and open the mask editing step.

text-to-image

Modal

Submit prompt and mask data for an active inpaint task.

text-to-image

Remix Strong

Remix a selected image with a stronger prompt change.

text-to-image

Remix Subtle

Remix a selected image with a lighter prompt change.

text-to-image

Parameters

ParameterTypeDefaultValuesNotes
versionselect8.28.2 | 8.1 | 8 | 7 | 6.1 | 6 | 5.2Model version. Prompt syntax and look differ between versions; older prompts often need rework on newer versions.
aspect_ratioselect1:11:1 | 16:9 | 9:16 | 4:3 | 3:4 | 3:2 | 2:3 | 21:9Output frame shape. Keep one ratio across a series.
speedselectfastrelax | fast | turboTurbo returns faster results and costs more.
qualityselect10.25 | 0.5 | 1 | 2Render quality/time trade-off. Higher q spends more render time.
stylizeslider1000–1000Controls artistic style intensity.
chaosslider00–100Higher values produce more varied results.
weirdslider00–3000Adds unusual composition and style choices.
rawtogglefalse—Reduces default aesthetic styling for more literal results.
tiletogglefalse—Creates seamless repeating patterns when supported by the selected version.
nijitogglefalse—Uses Niji anime styling. Select v7 or v6 for best compatibility.
seedtext–20Optional integer seed for repeatable results.
negative_prompttext–1000Things to avoid in the image.
sreftext–1000Optional URL used as a style reference.
creftext–1000Optional URL used as a character reference.
iwslider10–3Controls how strongly uploaded image references influence the result.

Pricing

Pricing will be published shortly — contact us for early access rates

Prices last updated: —

API example

POST /v1/generations — midjourney
curl -X POST https://api.mukelabs.com/v1/generations \
  -H "Authorization: Bearer $MUKE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "toolId": "midjourney",
    "inputMode": "text-to-image",
    "prompt": "your prompt here",
    "params": { "version": "8.2" },
    "references": [],
    "idempotencyKey": "a-unique-key-per-task",
    "subjectId": "your-user-id"
  }'

FAQ

How do I call Midjourney on Muke Labs?
Submit POST /v1/generations with toolId "midjourney" and one of its input modes (text-to-image, image-to-image), then poll GET /v1/generations/{jobId} for the result.
What does Midjourney cost?
Public pricing for this model is being finalized — contact us for early access rates. All tasks are billed from your prepaid USD balance and failed tasks refund automatically.
What happens if a task fails?
The job returns status "failed" with a typed error code, and its charge is returned to your balance automatically.
Where do I get an API key?
Create one in the developer portal (https://portal.muke.name), then send it as a Bearer token on every request.

Related models