Midjourney launches its first AI Video Generation Model, V1

Midjourney has released its first AI video generation model, V1, expanding its platform beyond AI image creation for the first time. V1 enables users to generate four 5-second video clips from one uploaded or previously generated image, although access is currently limited to web use on Discord.

V1 offers users the choice to let the model animate automatically or provide specific animation descriptions using text prompts. Motion intensity can be fine-tuned with 'low' and 'high' options, and individual videos can be extended in 4-second increments for a maximum total of 21 seconds. Midjourney includes video generation in its existing subscription tiers, starting at $10/month for the Basic plan. Pro ($60) and Mega ($120) plans offer unlimited access in "Relax" mode, which processes generations at lower priority but without credit limits. Each video costs 8x more credits than an image, though overall pricing remains slightly lower than alternatives like Google Veo 3.

With this debut, Midjourney enters the AI video space alongside Sora, Runway, Adobe Firefly, Google Veo, or the recently launched Seedance by ByteDance. CEO David Holz also mentioned plans to add audio generation, along with longer-term goals around 3D rendering and real-time simulation models