Midjourney Launches First AI Video Generation Model V1

Midjourney has emerged as a leading AI image generation startup, announcing on Wednesday the launch of its highly anticipated AI video generation model, V1.

V1 functions as an image-to-video tool, enabling users to upload an image or utilize images generated by Midjourney’s other models to produce four distinct five-second videos. Similar to Midjourney’s existing image models, access to V1 is exclusively through Discord and currently available only online.

The introduction of V1 positions Midjourney against established competitors in the AI video generation space, including OpenAI’s Sora, Runway’s Gen 4, Adobe’s Firefly, and Google’s Veo 3. While many companies focus on developing controllable AI video models for commercial applications, Midjourney has distinguished itself through its creative-focused AI image models.

According to the company, V1’s purpose extends beyond creating B-roll content for Hollywood films or advertising. In a blog post, CEO David Holz emphasized that the video model represents a strategic step toward building “real-time open-world simulations.”

Following the video model’s development, Midjourney plans to create AI systems capable of generating 3D renderings and real-time AI-driven models.

Notably, the release of V1 occurred just one week after Midjourney faced legal action from two major Hollywood studios—Disney and Universal. The lawsuit claims the startup’s AI-generated images depict copyrighted characters such as Homer Simpson and Darth Vader.

Hollywood studios are actively addressing the growing prevalence of AI tools like Midjourney’s, as concerns mount over potential impacts on creative industries. Several media companies allege these models were trained using their copyrighted materials.

While Midjourney positions itself as distinct from other AI startups by prioritizing creative applications over direct commercial use, it remains unable to escape these legal challenges.

Currently, video generation costs are eight times higher than standard image creation, meaning subscribers will exhaust their monthly quota more rapidly when producing videos.

The most affordable entry point for testing V1 is the $10/month basic plan. Subscribers to the $60/month Pro plan and $120/month Super plan gain unlimited video generation in the company’s slower “chill” mode. Midjourney plans to reassess V1’s pricing structure within the next month.

V1 incorporates customizable settings to control output quality. Users can choose between automated animation (which applies random movements) or manual settings with text-based instructions for specific animations. Movement intensity for both camera and subject motion can be adjusted via “low motion” or “high motion” options.

Though initial videos are limited to five seconds, users may extend content in four-second increments up to four times, resulting in a maximum 21-second video.

Early demonstrations of V1 videos exhibit a surreal yet visually compelling aesthetic, similar to Midjourney’s image outputs. While early feedback remains positive, comparisons against established models with months or years of market presence have yet to be fully evaluated.