As a generative image creation tool, Midjourney is perhaps best known for its operation within Discord servers, and now it is expanding its field of artificial intelligence. The creators of Midjourney announced on Tuesday that they plan to launch a "text-to-video" model in the coming months.

The company will begin training its video model starting in January, said CEO David Holz during an "office hours" Discord meeting. This move is a natural progression for the platform, which is based on mature image models to inspire competition in the generative video industry.

The meeting notes include plans to adjust V6 Niji, Midjourney's comic/animation generation model, and address consistency issues for the upcoming official release of Midjourney V6. The company also stated that its to-do list includes "starting training for a new video model," which could be ready "in a few months."

Holz and the Midjourney team have not shared more information about the model.

Midjourney is known for emphasizing quality and user experience rather than pursuing speed, even if it means lagging behind competitors. After several months of other platforms like Stable Diffusion turning internal filling and external expansion into reality, the company launched enhanced features, with recent forays into primary text generation following the common capabilities of models such as Dall-E 3, SDXL, and even less popular generators like Ideogram or IF.

Entering a crowded field

After competitors released new products, Midjourney also ventured into the field of video. Stability AI recently announced Stable Video Diffusion; Meta just showcased its EMU video generator, and existing models like Pika and Runway ML are also calibrating their domains, creating a stable competitive landscape for Midjourney. Additionally, other image generators like Leonardo AI have already achieved video generation capabilities, further intensifying the competition.

The recent v6 update from Midjourney, with improved prompt following capabilities and more realistic images, is the company's latest effort to maintain relevance and competitiveness. If its models demonstrate some coherence, they may even establish a solid foundation in such a nascent field, although the models are still far from perfect.

The impact of these developments extends beyond the power struggle between companies. As Midjourney and other companies innovate and refine their products, the creative and media industries are on the cusp of a transformative era. The ability to generate, manipulate, and interact with video content through artificial intelligence opens up many possibilities - from simplifying work for entertainers and advertisers to potentially reshaping our perception of reality.