Midjourney announces V7 image model with higher accuracy and Draft Mode

Midjourney has introduced an alpha version of its V7 image model, which produces images of noticeably higher quality.

The AI company calls their V7 image model its “smartest, most beautiful, most coherent model yet” that offers better details and more realistic textures, bodies, and hands than any of its predecessors.

The next-generation image creator introduces a brand-new feature called Draft Mode. In Draft Mode, images are generated in lower quality at half the cost, but according to Midjourney, they are rendered at 10 times the speed.

Users can make changes to their AI-generated images by changing the prompt, or they can enable Voice Mode by tapping the microphone button. “Tell it to swap out a cat with an owl or make it night time and it will automatically manipulate the prompt and start a new job,” Midjourney’s announcement of its V7 image model reads.

If users are happy with the results but would like to increase the quality of their image, they’ll only have to press ‘Enhance’ or ‘Vary’ on the image and the V7 image model will re-render the image at full quality.

V7 is the first model to have model personalization turned on by default. Users will have to specify in advance what pairs of photos they like best. They have to do this two hundred times, which takes about five minutes of their time to accomplish. Users can always decide to disable this feature afterwards.

Midjourney stresses that V7 is still in the alpha phase and is currently being tested by the community. Features such as upscaling, editing, and retexture will currently fall back to V6 models. However, Midjourney promises to release a new V7 upscaler soon for higher resolution.

“Expect new features every week or two so for the next 60 days. The biggest incoming feature will be a new V7 character and object reference,” the AI company teases. It also says that the new model has its strengths and weaknesses and recommends users to “play around a bit”.