muapi Logomuapi Logo
muapi Logomuapi Logo

Midjourney V7: AI Image Generator

Generate 4 photorealistic images per run with Midjourney V7. Supports text-to-image and reference image guidance via sourceimageurl. Midjourney V7 generates 4 high-quality photorealistic images per run from a text prompt or reference image. It combines advanced style control with sharp composition and vivid detail, making it ideal for everything from product visualization to creative concept art. Each generation produces a set of 4 varied images at $0.10 per run.

Interactive model controls

Generate 4 photorealistic images per run with Midjourney V7. Supports text-to-image and reference image guidance via source_image_url.

📝

Overview

About this model

Midjourney V7 generates 4 high-quality photorealistic images per run from a text prompt or reference image. It combines advanced style control with sharp composition and vivid detail, making it ideal for everything from product visualization to creative concept art. Each generation produces a set of 4 varied images at $0.10 per run.

1Photography: Generate hyperrealistic product and portrait photography without a camera.
2Concept Art: Create detailed character and environment concepts for games and film.
3Marketing: Produce compelling visual assets for campaigns and social media.
4Design: Rapidly iterate on visual ideas with reference image guidance.
5Illustration: Generate stylized illustrations and artistic compositions.
💰

Pricing & Value

Cost analysis

muapiapp/bin/zsh.10 per run (4 images)

Includes 4 upscaled images per generation with full parameter control.

Fal.aiNot available

Midjourney is not available on Fal.ai.

ReplicateNot available

Midjourney is not available on Replicate.

** Competitor pricing is estimated based on similar model architectures and usage tiers.

⚙️

Technical Details

Configuration schema

Promptstring

Text description of the image to generate.

Default ValueA forgotten royal bathhouse hidden deep inside sandstone cliffs during monsoon rain, warm candlelight reflecting across flooded marble floors, silk curtains moving gently with humid wind, subtle human presence, atmospheric depth, breathtaking architectural detail, timeless cinematic realism, premium editorial composition
Reference Image URLstring

Optional reference image URL. Influences the style and content of the generation.

Default Valueundefined
Aspect RatioEnum (6 options)

Output image aspect ratio.

Default Value1:1
Stylizeint

Controls how artistic the result is. Range: 0–1000. Lower = more literal, higher = more stylized.

Default Value100
Chaosint

Controls variation between the 4 images. Range: 0–100. Higher = more diverse.

Default Value0
Weirdint

Adds unconventional aesthetics. Range: 0–3000.

Default Value0
Negative Promptstring

Things to exclude from the image, e.g. "text, watermark".

Default Valuetext, watermark, blurry
Seedint

Reproducibility seed. Range: 0–4294967295. Same seed + same prompt ≈ similar result.

Default Value0
📖

Implementation Guide

Developer documentation

How to Use Midjourney V7

  1. Write a prompt: Be concise and evocative — use style keywords like "cinematic lighting", "hyperrealistic", "studio photography", or "oil painting".

  2. Add a reference image (optional): Provide a to guide the style or subject of the generation.

  3. Tune the parameters:

    • (0–1000): higher = more artistic interpretation
    • (0–100): higher = more variation between the 4 images
    • (0–3000): higher = more unconventional aesthetics
    • : things to exclude, e.g. "text, watermark"
    • : fix for reproducible results

  4. Pick an aspect ratio: , , , , , or .

  5. Poll for results: Each run returns 4 image URLs once complete (typically 30–90 seconds).

❓

Common Questions

Frequently asked

How many images does each run produce?

Every run generates 4 upscaled images. All 4 are returned in the output array.

What is the difference between stylize, chaos, and weird?

Stylize controls artistic interpretation (0 = literal, 1000 = highly stylized). Chaos controls how different the 4 images are from each other. Weird adds unconventional or surreal aesthetics.

Can I use a reference image?

Yes — provide a URL in the field. The reference image influences the style and content of all 4 generated images.

How long does a generation take?

Typically 30–90 seconds. Use the webhook callback URL to be notified when the images are ready.