Midjourney Unveils V1 Video Model for Clips Ranging from 5 to 20 Seconds

On June 18, 2025, Midjourney launched its first video model, V1, available to its approximately 20 million subscribers. Users can animate still images, generating four 5-second clips extendable to 20 seconds. With two motion modes—low for subtle movements and high for larger shifts—users can opt for automatic or text-prompted motion. Each clip costs around eight image credits, with a starting price of $10 per month. Although V1 lacks audio and is capped at 1080p, it is competitively priced compared to rivals like Runway Gen-4 and OpenAI’s Sora. While testers commend its coherence, some note flickering issues in high-motion scenes. Reactions vary, with some praising its performance and others highlighting limitations compared to higher-end models. The San Francisco-based lab, which generated $300 million from its image service in 2024, is currently facing a copyright lawsuit related to its training data. V1 is seen as an initial step toward more advanced capabilities.

Source link