Midjourney drops Omni-Reference for pinpoint image control

Midjourney just launched a new feature called Omni-Reference, making visual consistency across generations a whole lot easier.

According to the team, the tool gives creators a direct way to say "put THIS in my image." Instead of just matching a character's face, Omni-Reference works across characters, objects, vehicles, and other visual elements.

Beyond character consistency

Maintaining consistent props, vehicles, and subjects has long been one of the biggest headaches in AI image generation. Users typically had to rely on complex prompting or pure luck to keep specific items looking the same across different scenes.

With Omni-Reference now live, users can directly anchor specific characters, items, and vehicles into new prompts, expanding Midjourney's reference capabilities beyond standard portrait matching.

Why it matters

For digital artists, storyboard creators, and concept designers, consistency is the difference between a fun toy and a professional workflow. Getting the same character into a new scene is useful, but keeping their specific car, weapon, or signature prop identical is what makes serial storytelling actually viable.

Midjourney is smart to expand reference controls beyond just human faces. While the actual precision and flexibility will depend on real-world testing across complex scenes, Omni-Reference looks like a solid step toward making AI generation genuinely production-ready.