Midjourney, a prominent AI startup best known for image generation tools, unveiled its first-ever AI video creation model, V1, on Wednesday. This new technology allows users to upload an image—either one they’ve created themselves or previously generated by Midjourney’s own image models—and produce four distinct five-second videos based on that single input. At launch, users can access V1 exclusively through Discord and only via web platforms.
Midjourney’s distinctive style, aimed squarely at the creative community rather than purely commercial purposes, has always differentiated it from other AI providers in the marketplace. The arrival of V1 positions the startup directly against established rivals including OpenAI’s Sora, Runway’s Gen 4, Adobe’s Firefly, and Google’s Veo 3. However, the company insists that its long-term goal goes beyond simply competing for commercial video content such as promotional clips or cinematic B-roll.
In announcing V1, Midjourney CEO David Holz explained in a blog post that the new technology represents an important milestone in a broader journey towards offering AI models capable of creating “real-time open-world simulations.” Following video, Holz revealed Midjourney intends to pursue AI-generated 3D renderings and eventually real-time interactive environments.
Notably, V1’s announcement arrives at a controversial moment for Midjourney. Just a week ago, the startup became the target of litigation from entertainment industry heavyweights Disney and Universal. The lawsuit alleges copyright infringement related directly to characters such as Homer Simpson and Darth Vader appearing in images generated by Midjourney’s AI.
The entertainment sector is increasingly wary of AI-generated media due to concerns that such technologies might dilute or outright displace traditional creative jobs. Studios have regularly expressed concern over AI tools that allegedly rely upon proprietary intellectual property to train their algorithms—claims implicitly driving some of this industry’s pushback.
Despite the ongoing controversies, Midjourney’s video capabilities immediately attracted interest. For now, though, generating video comes at a premium: each video creation costs eight times as much as creating a single image, meaning subscribers’ monthly generation limits will reach their cap significantly quicker when producing videos. However, subscribers to Midjourney’s upper-tier plans ($60/month Pro plan and $120/month Mega plan) will receive unlimited video generations, provided they use the slower “Relax” mode.
Initially, users can experiment with V1 through the company’s lowest-priced Basic subscription at $10 per month. Midjourney acknowledged that pricing could change over the next month as they assess usage patterns and demand.
The V1 model offers several settings designed for creative flexibility. Users can choose an automatic animation feature that creates spontaneous movements, or manually specify detailed actions through text prompts. There are also settings to configure the degree of camera movement and subject motion, allowing for “low motion” or “high motion” outputs. Though initial V1 videos are limited to a five-second duration, users can extend these clips incrementally by four seconds up to four times, providing the potential for clips as long as 21 seconds.
Just as with Midjourney’s image generation models, early demonstrations of V1 video generation suggest a distinctive and somewhat surreal aesthetic, rather than aiming for hyper-realistic renderings. Initial reactions emphasize enthusiasm, although V1’s longer-term reception and competitiveness compared to its more mature rivals remain to be seen.
