Midjourney, the AI image generation startup best known for its imaginative visuals, has launched its first video generation model, V1, in a move that places the company in direct competition with industry giants such as OpenAI, Adobe, Google, and Runway.
Announced on Wednesday, V1 allows users to transform a single image — whether user-uploaded or generated via Midjourney’s image models — into a set of four five-second AI-generated videos. The feature is initially available via Discord, maintaining Midjourney’s longstanding platform approach, and can only be accessed on the web for now.
Unlike models focused on commercial-grade content such as OpenAI’s Sora or Google’s Veo, Midjourney’s V1 continues the startup’s creative-first philosophy. Its early outputs, as seen in user demos, are stylised and otherworldly rather than photorealistic, underscoring its artistic rather than industrial ambitions.
In a blog post accompanying the launch, CEO David Holz described V1 as part of Midjourney’s broader effort to develop tools for “real-time open-world simulations.” He further revealed that future plans include AI models for 3D rendering and interactive content generation, potentially signalling a pivot toward gaming or metaverse applications.
However, the launch comes at a contentious time. Just last week, Midjourney was sued by two major Hollywood studios — Disney and Universal — over the alleged use of copyrighted characters in its AI-generated imagery. The suit follows growing industry unease over how these AI systems are trained and what role they will play in the future of creative work.
While Midjourney has tried to distinguish itself from rivals with its emphasis on artistry and individual expression, the litigation underscores the legal risks AI startups face as they scale.
V1 also introduces new usage policies and pricing. Video generations cost roughly eight times more than image generations, meaning users will deplete their subscription credits significantly faster. The company’s $10 Basic plan offers limited access to V1, while higher-tier subscribers — those paying $60 or $120 a month — receive unlimited video access in a slower “Relax” mode.
To help creators tailor their results, V1 includes custom settings. Users can choose between automatic random animations or manually describe desired movements using text prompts. Motion levels can also be adjusted to “low” or “high,” and videos can be extended by four seconds at a time, up to a maximum of 21 seconds per clip.
The model’s release adds further competition to an already crowded and rapidly evolving field. Midjourney joins OpenAI’s Sora, Adobe’s Firefly, Google’s Veo, and Runway’s Gen-4, all of which are racing to develop controllable, high-fidelity AI video tools.
Despite its limited control options and abstract visuals, early feedback on V1 has been broadly positive. Analysts say its creative orientation may appeal to a niche yet loyal segment of digital artists and storytellers.
