AI Usage at a Glance
Jan 4, 2024
OtherPractice documented: Midjourney trains its AI image and video generation models on datasets that include billions of images crawled from the public web, licensed data, and user-submitted prompts and outputs. This training practice is active and ongoing since 2022.
Practice DocumentedView practice →Aug 16, 2024
Creative GenPractice documented: Midjourney deployed a web-based image editor in August 2024 that lets users modify AI-generated or uploaded images using AI-powered tools. Users can repaint specific areas of an image using text prompts (inpainting) or expand an image's canvas beyond its original borders (outpainting).
Practice DocumentedView practice →Aug 21, 2024
Creative GenNew evidence: Midjourney Unveils Unified Web-Based AI Image Editor
Evidence AddedView practice →Jan 1, 2025
ModerationPractice documented: Midjourney uses automated filters that scan user text prompts and generated images to block content that violates its Community Guidelines, including explicit, violent, and harmful material. These filters apply to all users on all plans across all platforms at all times.
Practice DocumentedView practice →Apr 4, 2025
RecommendationPractice documented: Midjourney offers a Personalization feature, introduced as a default element of V7 (April 2025), that learns a user's visual preferences by having them rate images and then subtly adjusts future AI-generated images to match their individual aesthetic taste. Users can turn this feature on or off at any time.
Practice DocumentedView practice →May 1, 2025
Creative GenPractice documented: Midjourney deployed an AI image generator that creates original artwork from text descriptions typed by users. Since its open beta launch in July 2022, any paying subscriber can type a written prompt and receive four AI-generated images within about a minute.
Practice DocumentedView practice →Jun 11, 2025
OtherNew evidence: How the Disney-Midjourney Suit Could Reshape AI Copyright Law
Evidence AddedView practice →Jun 18, 2025
Creative GenPractice documented: Midjourney deployed an image-to-video generation feature (V1 Video Model) in June 2025 that allows paying subscribers to animate any still image into a short video clip. Users select an existing or uploaded image and the AI produces a five-second video that can be extended up to 21 seconds.
Practice DocumentedView practice →Jun 19, 2025
Creative GenNew evidence: Midjourney adds AI video generation
Evidence AddedView practice →Mar 18, 2026
Creative GenPractice documented: Midjourney offers a specialized image generation model called Niji, developed in collaboration with Spellbrush, that focuses on Eastern and anime illustration aesthetics. It is available to all paying Midjourney subscribers and has its own dedicated website and Discord server.
Practice DocumentedView practice →Midjourney deployed an AI image generator that creates original artwork from text descriptions typed by users. Since its open beta launch in July 2022, any paying subscriber can type a written prompt and receive four AI-generated images within about a minute.
Midjourney's core product is a generative AI model that accepts text prompts (and optionally reference images) as input and produces sets of four images as output. The model has been updated through multiple versions — V6 (December 2023), V6.1 (July 2024), V7 (April 2025, became default June 2025), and V8.1 (alpha, 2026) — each improving prompt accuracy, image coherence, and generation speed. The service is accessible via Discord bot commands and the midjourney.com web interface. Subscription plans range from $10 to $120 per month; no free tier has been available since 2023. The platform has over 20 million registered users and reportedly generates approximately 2 billion images annually.
Midjourney deployed an AI image generator that creates original artwork from text descriptions typed by users. Since its open beta launch in July 2022, any paying subscriber can type a written prompt and receive four AI-generated images within about a minute.
Midjourney offers a specialized image generation model called Niji, developed in collaboration with Spellbrush, that focuses on Eastern and anime illustration aesthetics. It is available to all paying Midjourney subscribers and has its own dedicated website and Discord server.
Midjourney deployed a web-based image editor in August 2024 that lets users modify AI-generated or uploaded images using AI-powered tools. Users can repaint specific areas of an image using text prompts (inpainting) or expand an image's canvas beyond its original borders (outpainting).
Midjourney deployed an image-to-video generation feature (V1 Video Model) in June 2025 that allows paying subscribers to animate any still image into a short video clip. Users select an existing or uploaded image and the AI produces a five-second video that can be extended up to 21 seconds.
Have evidence about Midjourney's AI practices? Submit a report.
Submit a report →AI Trace is free and nonprofit. Support our work
Midjourney offers a specialized image generation model called Niji, developed in collaboration with Spellbrush, that focuses on Eastern and anime illustration aesthetics. It is available to all paying Midjourney subscribers and has its own dedicated website and Discord server.
Niji is a series of AI image generation models optimized for anime, manga, and Eastern illustration styles. Developed in partnership with Spellbrush, the models are tailored to understand anime conventions, character archetypes, and Japanese animation visual storytelling. Niji 7 was launched on January 9, 2026, with improvements in coherency and fine detail rendering. Niji models are available via the same Discord bot interface and Midjourney website used for standard models, though Spellbrush also operates a dedicated website (nijijourney.com) and Discord server for the Niji community.
Midjourney trains its AI image and video generation models on datasets that include billions of images crawled from the public web, licensed data, and user-submitted prompts and outputs. This training practice is active and ongoing since 2022.
According to Midjourney's AB2013 Documentation (a California regulatory disclosure), its models are trained on datasets comprising billions of images, text, and audiovisual content from multiple source categories: publicly crawled web content, licensed data, public domain data, and data potentially protected by copyright used under a fair use claim. User-submitted prompts and generated outputs are also covered by Midjourney's Terms of Service, which grants the company a perpetual license to use them for service improvement. Training data undergoes processing steps including safety filtering (to remove CSAM and other sensitive content) and privacy processing to filter personal information. As of 2025, Midjourney faces multiple active copyright infringement lawsuits — including a class action by artists (Andersen et al., filed January 2023) and a lawsuit filed by Disney and Universal in June 2025 — challenging whether its training on copyrighted works constitutes fair use.
