Midjourney Introduces V7 with Accelerated Generation and Personalization
Midjourney has announced the seventh version of its neural network for creating images. The main innovations are a fast draft generation mode (Draft Mode) and built-in personalization, allowing the model to better understand user requests.

Midjourney has released the seventh version of its neural network for generating images, V7. Key innovations include improved quality, photorealism, and anatomy, as well as two key features: Draft Mode and default personalization.
Draft mode allows you to generate images 10 times faster and twice as cheaply using voice input in a dialog format. The user can speak out ideas, and the model will quickly correct the sketches. To activate, just add —draft to the request or use the microphone.
Finished sketches can be further developed in the standard mode. Personalization in V7 is enabled by default: after a 5-minute survey, the model adapts better to the user's preferences. Detail, textures and object consistency have also been improved.
There are two modes of operation available - Turbo (fast but expensive generation) and Relax (standard speed). For now, scaling and editing work on V6, but will be updated in the future.
The team plans to release new features every 1-2 weeks for two months. Users can test the alpha version of V7 and leave feedback in the official Midjourney channels.
What's Your Reaction?






