Midjourney reveals what's next for v6 and video

If you jumped to Midjourney v6 and missed your favorite editing buttons, relief is around the corner.

In the January 17 Office Hours call, founder David Holz confirmed that zoom, pan, and vary region are set to arrive on v6 as early as next week.

Feature parity and a push toward video

The team is preparing to move v6 from beta to alpha, backed by more fine-tuning for the base model.

Consistency features are also rolling out in stages: style consistency will land before character consistency. On top of that, v6 will get a rebuilt style tuner that works differently from the v5 version and remains distinct from style consistency.

Outside of Discord, Midjourney plans to expand access to its alpha website and completely overhaul the explore page.

Looking further ahead, video generation is actively in development and will debut before 3D features. Further down the pipeline, v7 is expected to bring a massive leap in prompt accuracy.

David also shared a simple breakdown of the stylize parameter: lower values force the model to follow your prompt strictly, while higher values let the AI do whatever it thinks looks best.

Why it matters

For daily creators, getting inpainting and panning back on v6 means real production work can finally resume. High fidelity means little if you cannot fix a bad hand or extend a canvas.

Tackling style consistency before character consistency looks like a smart tactical move. Locking down color palettes, lighting, and rendering styles is technically easier than nailing facial consistency, yet it immediately solves a massive pain point for commercial designers and brand work.

Meanwhile, expanding the web alpha is critical. As rivals build standalone creative suites, Midjourney needs to prove it can outgrow Discord and deliver video before dedicated tools lock up that market.