AI firm Midjourney is now starting the “alpha-test phase” of V7, its new AI image-generation model. The V7 model is now available on both the Midjourney website and the platform’s Discord-based chatbot interface.
“We’re now beginning the alpha-test phase of our new V7 image Model. It’s our smartest, most beautiful, most coherent model yet. Give it a shot and expect updates every week or two for the next two months,” Midjourney announced in a post on X. According to founder and CEO David Holz, the new new model is “much smarter with text prompts,” and that the “image quality is noticeably higher with beautiful textures, and bodies, hands and objects of all kinds have significantly better coherence on all details.”
We’re now beginning the alpha-test phase of our new V7 image Model. It’s our smartest, most beautiful, most coherent model yet. Give it a shot and expect updates every week or two for the next two months. pic.twitter.com/Ogqt0fgiY7
— Midjourney (@midjourney) April 4, 2025
Midjourney is framing the current V7 release as a community-driven alpha test. Users are encouraged to experiment and provide feedback on the new model’s strengths and limitations. Furthermore, Midjourney has outlined a biweekly release schedule for new features over the next two months, and plans to bring features such as a character and object reference system, which would allow for more controlled and consistent representations across multiple image generations.
The alpha version of the model is set to come with a default personalization mechanism. To activate it, users are required to complete a brief onboarding process involving the rating of approximately 200 image pairs. This setup enables the model to align with the user’s visual preferences, creating a custom-tailored image-generation experience. Personalization does not consider the semantic content of prompts, but instead trains the model to recognize and emulate the styles, aesthetics, and visual traits a user prefers. This allows the system to produce outputs that are more aligned with individual taste, albeit requiring some upfront user interaction to unlock the feature.
With V7, Midjourney also launched a new feature called Draft Mode. This function enables the rendering of images at a faster rate (by 10 times, according to the official statement), while consuming only half the computational credits. The trade-off is a reduction in image quality, but the platform offers an “enhance” function that lets users re-render any draft output at full quality if desired. In addition to this, the Draft Mode introduces a conversational prompt interface on the web version of Midjourney. In this mode, users can give real-time feedback or suggest alterations without rephrasing the entire prompt. Voice commands are also supported through a “voice mode” function, accessible after enabling Draft Mode.
V7 supports two operational modes — Relax and Turbo. Turbo mode, as the name implies, is optimized for faster rendering times but incurs higher operational costs, roughly twice the credits of a standard image generation job under the previous V6 system. Relax mode provides a slower but more economical alternative. A standard speed mode for V7 is under development and expected in future updates. While Draft and Turbo modes support rapid prototyping and idea iteration, several key features such as upscaling, inpainting, and retexturing are not yet available in V7 and currently revert to V6.1 for execution, though these capabilities will be integrated into V7 in subsequent releases.
Content originally published on The Tech Media – Global technology news, latest gadget news and breaking tech news.
