Midjourney has launched the alpha version of its V7 AI Image generation model amid the boom in the AI-powered image generator. The new version is expected to provide better image quality and improved text prompt comprehension.
What makes it special is its capability to render complex elements like hands, which have been a big challenge for most AI image generators.
Midjourney’s Entirely New Model ‘V7’
According to the Midjourney team, V7 is an “entirely new” model designed to produce higher fidelity images with better textures and anatomical accuracy. Early adopters have shared photorealistic samples of human hands on social media, which show the model’s refined capabilities. While still in its alpha phase, Midjourney plans to roll out incremental updates every one to two weeks over the next two months.
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
One of the main features of V7 is the introduction of “Draft Mode,” which is a cost-effective option that generates images at half the usual rendering time. However, this may reduce detail.
This feature is mainly useful for brainstorming and rapid iteration. Users can use voice input for easy collaboration, which eliminates the need for typed prompts. If a draft image shows promise, options to “enhance” or “vary” it allow for a high-resolution refinement.
At launch, V7 will include Turbo and Relax modes for standard rendering. Turbo mode offers faster processing, but at double the cost of a standard V6 image generation. However, standard-speed rendering remains under optimization and will be integrated in future updates. Certain features, such as upscaling, editing, and retexturing, are still dependent on V6 but will gradually shift to V7 as development progresses.
Testing the alpha version requires users to activate a “Global Personalization Profile” by ranking 200 image pairs. This feature can be enabled by default but can be deactivated if desired.
Source: https://x.com/midjourney/status/1908012961840672947
Latest Stories:
