Midjourney Unveils Alpha Release of V7 Image Generation Model with Major Upgrades
Midjourney has officially launched the alpha version of its V7 image generation model, now open for testing by the AI community. This latest release introduces significant advancements in prompt comprehension, image clarity, and feature consistency.
“V7 is an amazing model,” Midjourney shared. “It’s much smarter with text prompts, image prompts look fantastic, image quality is noticeably higher with beautiful textures, and bodies, hands, and objects of all kinds have significantly better coherence on all details.”
Smarter, Sharper, and More Intuitive
One of the standout innovations in V7 is the default activation of model personalisation. While users need to unlock this feature (a quick five-minute setup), it can be turned on or off at any time. Once enabled, personalisation helps the AI better interpret your preferences and creative style, giving a more tailored experience. Midjourney sees this as a leap forward in understanding user intent.
Meet Draft Mode: Faster, Cheaper, and Conversational
Alongside V7, Midjourney is rolling out a brand-new feature called Draft Mode. Designed to generate images 10x fasterand at half the cost, this mode allows for rapid prototyping of ideas.
Thanks to Draft Mode’s speed, Midjourney is introducing a conversational interface on its web platform. Now, users can simply ask the AI to tweak images—like swapping a cat for an owl or changing a sunny scene to night—and the system will auto-adjust the prompt and regenerate the image on the fly.
Adding to the convenience, voice input is also supported. With just a tap of the microphone icon, users can speak their ideas aloud and watch as the model brings them to life in near real-time.
While images in Draft Mode are lower in quality, their stylistic consistency remains strong. If a draft image hits the mark, users can upgrade it with the ‘enhance’ or ‘vary’ options for a full-quality render.
Speed Tiers, Cost Structure, and What’s Coming Next
Initially, V7 will support two speed modes: Turbo and Relax. The standard speed mode is still being fine-tuned and will roll out soon. Turbo mode jobs cost twice as much as standard ones, while Draft Mode jobs are half the cost.
Some existing features like upscaling, editing, and retexturing will temporarily fall back to using the V6 model, but upgrades are on the roadmap. Tools like mood boards and SREF are already functional and will see performance boosts in upcoming updates.
Midjourney has also laid out an ambitious development timeline, promising new features every 1–2 weeks for the next two months. One highly anticipated addition is the V7 character and object reference system, which will help further refine image control and consistency.
A Fresh Start with V7
Midjourney emphasizes that V7 is a brand-new model, with its own unique capabilities and quirks. They encourage users to explore, experiment, and share feedback, noting that prompting strategies may need to be adjusted compared to previous versions.