Midjourney Captioned Full: 23 Million AI-Generated Images and Prompts

23,167,456 Midjourney-generated images and captions were compiled by deepghs and an anonymous provider as of December 2024. This collection contains original image files alongside metadata such as dimensions and unique identifiers.

Use Cases

• Training text-to-image models using the text prompts and corresponding image files
• Benchmarking zero-shot image classification by comparing model outputs against the provided captions
• Analyzing aspect ratio distributions in AI art using the width and height columns

Strengths

• Scale of 23,167,456 records
• WebDataset format for high-throughput data loading
• Preservation of original image files rather than external URLs

Limitations

• Presence of not-for-all-audiences (NSFW) content
• Captions are raw user prompts which often contain non-descriptive stylistic keywords
• Anonymous source prevents verification of exact scraping or filtering methodologies

Provenance

Source

Midjourney

Collection Method

Scraped

Freshness

Last updated December 2024.