TLDR
- Midjourney launched its first AI video generation model V1, allowing users to create 5-second videos from images
- The service costs 8x more than image generation but starts at $10/month, competing with rivals like Runway and OpenAI’s Sora
- V1 produces four video options per generation and can extend clips up to 20 seconds total
- The launch comes days after Disney and Universal sued Midjourney for copyright infringement over AI-generated content
- Midjourney aims to eventually build “world models” for real-time interactive simulations beyond simple video generation
Midjourney launched its first AI video generation model V1 on Wednesday, entering the competitive AI video market. The tool transforms static images into 5-second video clips through the company’s Discord platform and website.
The image-to-video model allows users to upload existing images or use Midjourney-generated content as source material. Each generation produces four different video options that users can extend up to 20 seconds total.
V1 offers two main animation modes for users. The automatic setting creates random movement from static images, while manual mode lets users write text descriptions for specific animations they want to see.
Users can control motion intensity through low and high motion settings. Low motion creates subtle effects like blinking or gentle scene movement, while high motion attempts more dynamic camera and subject animation.
The pricing structure charges 8x more for video generation compared to standard image creation. Basic subscriptions start at $10 per month, with Pro and Mega subscribers getting unlimited video generations in “Relax” mode.
Midjourney’s entry puts it in direct competition with established AI video platforms. OpenAI’s Sora starts at $20 monthly, while Runway’s plans begin at $12 monthly and Google’s Veo 3 costs $249.99 monthly.
The model currently lacks audio generation capabilities that competitors offer. Users must add soundtracks separately in post-production, unlike services like Luma Labs’ Dream Machine which generates accompanying audio.
Copyright Lawsuit Clouds Launch
The video model launch comes just days after Disney and Universal filed a major copyright lawsuit against Midjourney. The studios accuse the company of training AI models on copyrighted characters without permission.
The complaint spans over 100 pages and names characters from Marvel, Star Wars, The Simpsons, and Shrek. Disney and Universal claim Midjourney creates a “bottomless pit of plagiarism” by allowing users to generate derivative content.
The lawsuit preemptively targets Midjourney’s video service, claiming the company was already training the model on protected IP before launch. Studios argue this gives Midjourney commercial advantage built on uncompensated creative work.
Disney’s general counsel Horacio Gutierrez stated that AI companies cannot escape copyright law. The case could set important precedents for how AI platforms handle training data and content filtering.
Future World Model Ambitions
Midjourney positions V1 as a stepping stone toward larger goals beyond video generation. CEO David Holz describes plans for “real-time open-world simulations” that users can navigate and explore.
The company envisions merging image generation, video, 3D rendering, and real-time interaction into unified world models. These systems would let users explore dynamically generated environments like interactive video games.
Midjourney faces competition from other companies pursuing similar world model technology. Odyssey streams 30fps video with spatial navigation, while Runway’s Gen-4 offers advanced camera controls for scene exploration.
The video model represents one component in Midjourney’s broader technical roadmap. The company says it must first build static visuals, motion, spatial control, and real-time responsiveness before achieving full world simulation.
Early user reactions to V1 have been positive, with creators praising the quality and pricing. However, the model’s 20-second limit and lack of audio generation leave room for improvement compared to competitors.
