On Friday, Midjourney launched its newest image-generation artificial intelligence (AI) model, V7, which is now available in an alpha version for user testing. This release comes nearly a year after the introduction of V6 and marks a significant advancement, as V7 is the first model from the company to have personalisation enabled by default. Alongside this, a new feature called “Draft Mode” has been introduced, allowing users to generate images at a speed ten times faster than the standard mode at half the cost.
Midjourney’s V7 Unveils Innovative Modes and Features
Midjourney announced V7’s release in a post on X (formerly Twitter). The new model demonstrates improved adherence to text prompts compared to V6, with enhanced image quality and texture, as well as better coherence among various elements within the images, according to the company.
One notable feature of V7 is its automatic personalisation, which requires users to rate a substantial number of images to establish a unique profile that informs the AI’s understanding of their preferences. This feature allows for customized image outputs but can be turned on or off by users after gaining access to the model.
Additionally, the introduction of Draft Mode is a significant enhancement. This mode, which is half the price of the standard mode, enables users to quickly iterate on ideas by providing multiple prompts and making inline edits to images. It also supports voice input, aiming to create a more conversational user experience.
One of the most exciting new features for our new V7 model is something we call “Draft Mode”. Draft mode is half the cost and 10 times the speed and it might be the best way to iterate on ideas ever. Try it with voice, think out loud and let our ideas flow like liquid dreams. pic.twitter.com/ANfTMC6Ej1
— Midjourney (@midjourney) April 4, 2025
To activate Draft Mode, users can either select it via the web client or append “–draft” to their prompt. Midjourney clarified that images generated in Draft Mode are of lower quality than those produced in the standard mode; however, the model’s behavior and aesthetic generation remain consistent.
The V7 model is accessible in both Turbo and Relax variants, with Turbo mode priced at twice the standard output level. It is important to note that V7 currently does not support upscaling, editing, or retexturing tasks, which will continue to redirect users to V6 for these functions.
Looking ahead, Midjourney intends to introduce new features for the V7 model bi-weekly over the next 60 days, including planned updates for character and object reference capabilities.
