V1 Video Model Overview
Midjourney has introduced its first video model, V1, expanding its AI capabilities from image generation to animated video creation. This new feature allows users to transform static images into dynamic video clips, catering to the growing demand for short-form content across digital platforms.
Image-to-Video Model Functionality
The V1 video model enables users to animate both images generated by Midjourney and external uploads. Users can select between automatic animation—where the AI determines movement—and manual mode, which lets users describe specific animations using text prompts. The video model supports two motion settings, offering flexibility in animation style.
5–21 Second Video-Model Clips
V1’s video model produces video clips ranging from 5 to 21 seconds. Videos start at 5 seconds and can be extended by 4 seconds up to four times, reaching a maximum of 21 seconds. This duration is ideal for social media, aligning with formats like YouTube Shorts and TikTok, and encourages creators to focus on impactful, loopable visuals.
Low vs High Motion Settings in the Video Model
The video model offers “Low motion” for subtle, steady animations and “High motion” for more dynamic, energetic clips. While “Low motion” keeps camera and subject movement minimal, “High motion” introduces significant movement, though it may sometimes affect output accuracy. Early feedback suggests the difference can be subtle, but these controls help tailor the video model’s output to creative needs.
Video Model Output and Access
Videos generated by the V1 video model are output as 480p MP4 files at 24 frames per second. There is currently no upscaling feature. The video model is accessible via Discord and the Midjourney website, with video generation costing 8x more GPU time than standard images.
For more on structuring content and best practices for headings, see Website Headings and Subheadings: Help Readers and SEO in 2025[1].
For Markdown link formatting, refer to The Dead Simple Markdown Guide to Links[2].
