Midjourney

Midjourney is a leading AI image generation platform known for artistic, high-quality outputs. It operates primarily through Discord and a web interface, using a command-based workflow.

Platform Overview

• Provider: Midjourney Inc.
• Specialty: High-quality artistic image generation
• Access: Subscription required (no free tier)
• Supported Output Types: Image

Lazer Detection

URL Patterns

The extension detects Midjourney on:

• https://www.midjourney.com
• https://midjourney.com/app
• Discord integration (limited support)

What Is Auto-Detected

When you generate with Midjourney, Lazer extracts:

• Prompt - From Discord message or web UI prompt display
• Model - From --v parameter (e.g., "v6.1", "v7")
• Aspect Ratio - From --ar parameter (e.g., "16:9", "1:1")
• Stylize - From --s parameter
• Chaos - From --c parameter
• Output URL - Image src from generation grid or upscaled result
• Thumbnail - Same as output URL

Detection Quality

• Prompt extraction: Medium - Works on web UI, limited on Discord
• Settings extraction: High - Parameters in prompt are parsed
• Output extraction: High - Reliable image URL detection

Capture Workflow

Web Interface Workflow

1. Open Midjourney web app
2. Enter prompt with parameters (e.g., /imagine a cat --v 6.1 --ar 16:9)
3. Wait for initial grid (4 variations)
4. Click upscale button (U1, U2, U3, or U4)
5. Wait for full-resolution image
6. Open Lazer extension
7. Detection banner shows "Midjourney"
8. Click "Auto Fill"
9. Review fields
10. Click "Save Capture"

Discord Workflow (Limited)

Detection on Discord is limited due to platform constraints:

1. Use /imagine command in Discord
2. Wait for grid generation
3. Upscale chosen variant
4. Right-click image and "Open Image in New Tab"
5. Tab shows cdn.midjourney.com URL
6. Open Lazer extension
7. Manually fill form (prompt, URL)
8. Click "Save Capture"

Parameter Parsing

Midjourney parameters are extracted automatically:

• --v 6.1 → Model: "v6.1"
• --ar 16:9 → Aspect Ratio: "16:9"
• --s 500 → Stylize: 500
• --c 20 → Chaos: 20
• --seed 123 → Seed: 123

All captured in metadata JSON.

Capture Tips

Best Practices

• Capture upscaled images - Initial grid is low-res
• Include parameters - Always use --v, --ar for consistency
• Note variations - U1-U4 indicate grid position
• Use seed for consistency - --seed enables reproducibility

Common Issues

Grid vs Upscaled:

• Midjourney generates 4-image grid first
• Upscaling produces full-res single image
• Capture upscaled version for production use

Discord detection limited:

• Discord's UI is complex and changes frequently
• Manual capture recommended on Discord
• Web interface has better detection support

Prompt with parameters:

• Full prompt includes parameters (e.g., --v 6.1)
• Detector parses parameters into metadata
• Original prompt is preserved without parameters

Rights and Licensing

Ownership by Plan

Midjourney licensing depends on your subscription tier:

Basic, Standard, Pro Plans

• Non-subscribers: Cannot use Midjourney
• Basic/Standard: Commercial license included if earning < $1M/year
• Pro/Mega: Full commercial license, no revenue cap

Commercial Use

• All paid plans include commercial use rights
• You own the outputs you generate
• Attribution not required but encouraged
• Subject to Midjourney Terms of Service

Restrictions

• Cannot use for illegal or harmful content
• Cannot use to train competing AI models
• Subject to content moderation and community guidelines

Lazer Rights State

The extension sets rightsState to:

• LICENSED - Default for authenticated Midjourney users (paid plans)
• UNKNOWN - If plan tier unclear

Manually verify your plan's commercial license terms on Midjourney's website.

Known Limitations

Discord Integration

• Discord's structure makes detection challenging
• Manual capture recommended for Discord workflow
• Web UI provides better auto-detection

Real-Time UI Changes

Midjourney frequently updates the web interface:

• Detector uses multiple fallback selectors
• Report detection failures with screenshots

Upscale Variants

Midjourney supports:

• U1-U4 (upscale specific grid position)
• Variations (V1-V4)
• Zoom out, pan, vary (region), etc.

Detector captures final output, not intermediate states.

Metadata Fields

Midjourney-specific metadata:

{
  "modelVersion": "v6.1",
  "aspectRatio": "16:9",
  "stylize": 500,
  "chaos": 20,
  "seed": 123456,
  "gridPosition": "U3",
  "platform": "midjourney"
}

Example Capture

A typical Midjourney capture in Lazer:

• Platform: midjourney
• Asset Type: IMAGE
• Prompt: "A cyberpunk cityscape at night, neon lights reflecting on wet streets"
• Model: v6.1
• Output URL: https://cdn.midjourney.com/abc123/grid_0.png
• Thumbnail: (same)
• Metadata: {"modelVersion": "v6.1", "aspectRatio": "16:9", "stylize": 500}
• Status: GENERATED
• Rights State: LICENSED

Troubleshooting

Detection Not Working

• Verify you're on the web UI (not Discord)
• Wait for upscaled image to fully load
• Click "Refresh Detection"
• Grant host permissions for midjourney.com

Prompt Extraction Failed

• Discord prompts may not be detected
• Manually copy prompt from Discord message
• Include all parameters in prompt field

Output URL Wrong

• Ensure you're capturing upscaled image, not grid
• Right-click image and "Copy image address"
• Manually paste URL if auto-detection fails

Parameters Not Parsed

• Verify parameters are in prompt (e.g., --v 6.1)
• Detector parses common parameters automatically
• Manually add to metadata if missed

Commercial License Unclear

• Review Midjourney subscription page
• Basic/Standard plans: < $1M/year revenue
• Pro/Mega plans: unlimited commercial use
• Consult Midjourney support if unsure

Related Resources

• Midjourney Homepage
• Midjourney Documentation
• Midjourney Subscription Plans
• Discord Server

Next Steps

• Compare with Freepik
• Understanding rights tracking
• Capture workflow guide