Midjourney is ready to make videos
Midjourney CEO David Holz just announced that the team will start training video generation models in January. On top of that, Midjourney v6 is getting an update next week to improve text rendering and prompt following, while gradually restoring v5 staples like inpainting. The reigning king of image generation is finally moving into motion.
Midjourney is officially ready to bring its visual muscle to motion. During a recent Discord "Office Hour" session, CEO David Holz announced that the company will begin training its video models in January.
Expanding beyond static imagery
Video isn't a total surprise, as Holz has previously hinted that Midjourney is developing both 3D and video capabilities.
Meanwhile, the recently released Midjourney v6 is already in line for improvements. Holz noted that updates rolling out next week will focus on better text rendering and prompt following, while popular v5 features like inpainting will gradually be restored to the new model.
Why it matters
Midjourney is widely considered the gold standard for image aesthetics, so watching it step into video is naturally exciting. That said, starting training in January means creators won't get their hands on a finished tool anytime soon, so there is no need to rethink your video stack just yet.
We think Midjourney could hold a real qualitative edge here by building on its mature image engine, much like Meta did when it derived the well-received Emu Video from its Emu image base. But motion demands consistent physics and immense compute over time, and whether Midjourney can replicate its image supremacy in video remains an open question.
Sources (1)
- the-decoder.com(opens in a new tab)Originalthe-decoder.com
Topic trackerMidjourney v6 and Video Development
- Midjourney drops Niji V6 for anime art
- Midjourney reveals what's next for v6 and video
- Midjourney is ready to make videosThis story
- Midjourney v6 is now live