Midjourney推出首个AI视频生成模型V1

Core Viewpoint - Midjourney has launched its first AI video generation model, V1, entering the competitive AI video space alongside companies like OpenAI and Runway [1] Group 1: Product Overview - The V1 model is designed as an image-to-video tool, allowing users to upload their own images or use images generated by other Midjourney models to create four five-second video clips [5] - Currently, the V1 functionality is available exclusively through the web version on the Discord platform, continuing Midjourney's previous product service path [5] - CEO David Holz stated that the launch of V1 is a crucial step towards the company's ultimate goal of "real-time open world simulation," rather than being limited to specific applications like Hollywood movie production or commercial advertising [5] Group 2: Technical and Pricing Details - Following the V1 model, Midjourney plans to develop 3D rendering AI models and real-time AI models, indicating a comprehensive technical layout in multimedia generation [6] - The V1 model employs a differentiated pricing strategy, with video generation costs being eight times that of typical image generation, requiring users to consume more monthly generation quotas [6] - The lowest tier for trying this feature starts at $10 per month for the Basic plan, while the Pro plan at $60 and the Mega plan at $120 offer unlimited video generation in "Relax" mode [6] - Users have various customization options, including automatic animation mode for random effects and manual mode for precise animation requests, with the ability to adjust movement speed [6] - Although the default video length is five seconds, users can extend it up to 21 seconds, and initial feedback on the generated videos has been positive, although its overall competitive strength needs further validation in practical applications [6]