Midjourney, one of the pioneers in AI-powered image generation, has launched its first new model in nearly a year—V7. The update, released in alpha early Thursday, comes at a pivotal moment in the AI art space, just one week after OpenAI’s DALL-E 3 integration in ChatGPT went viral for its ability to create stunning Studio Ghibli-style artwork.
The Midjourney V7 is not explicitly trained for a specific art style, but some initial experimentation shows it generates gorgeous detail and visual consistency across the broadest array of aesthetics. Among several major developments, this model will settle a few ancient woes in the realm of AI image generation: understanding prompts better, producing much finer-quality textures, and actually getting hands right-a notorious flaw in most AI art generators till this day.
How to Access Midjourney V7
Before using V7, users must first rate around 200 images to create a personalized profile—a new feature that tailors outputs to individual preferences. This training process helps the AI understand your aesthetic tastes, whether you prefer photorealistic images, painterly styles, or something more abstract.
Once set up, users can toggle V7 on or off in:
- Mid Journeys web app (via the “Version” dropdown menu)
- Mid Journeys Discord bot (for server members)
The model comes in two distinct modes:
- Turbo (prioritizes speed but consumes more GPU minutes)
- Relax (standard speed for more budget-conscious creation)
A particularly innovative addition is the new Draft Mode, which allows for 10x faster rendering at half the cost. While these initial outputs are lower quality, they serve as excellent starting points that can be refined and upscaled with a single click—perfect for rapid prototyping of ideas.
Key Improvements in V7
Midjourney CEO David Holz called V7 a “complete architectural overhaul” that represents the company’s most significant technological leap since its launch. The improvements are immediately noticeable:
Advanced prompt comprehension – The model now better understands complex, multi-concept prompts and maintains consistency throughout generations
- Unprecedented image quality – From fabric textures to metallic reflections, details pop with new realism
- Anatomical coherence – Fewer distorted hands, properly proportioned limbs, and more natural poses
- Style consistency – Maintains uniform aesthetics across multiple generated images
- Enhanced image prompting – Better at incorporating and blending reference images
However, some fan-favorite features—particularly upscaling and texturing tools—are temporarily unavailable in V7. Holz estimates these will return within two months as the team continues refining the model.
Testing V7: First Impressions
In hands-on testing, V7 demonstrates remarkable adherence to prompts while showing greater creative interpretation when given vague instructions. The model seems particularly adept at:
- Architectural visualizations with clean lines and proper perspective
- Fantasy character designs with intricate armor and clothing details
- Food photography that looks genuinely appetizing
- Abstract compositions with thoughtful color relationships
Holz encourages users to experiment with different prompting styles, noting that V7 may respond differently than previous versions. The community is already sharing prompt formulas that take advantage of V7’s enhanced capabilities.
Midjourney’s Unconventional Path
Founded in 2022 by David Holz (co-founder of the motion control company Leap Motion), Midjourney has charted a unique course in the AI industry. Unlike competitors backed by tech giants, the company has:
- Grown entirely without external funding
- Maintained a lean team of just 11 full-time staff
- Generated an estimated $200 million in revenue by late 2023
The company recently announced it’s assembling a hardware research team, signaling ambitions beyond software. While details remain scarce, this move suggests Midjourney may be developing specialized AI chips or other infrastructure to support its next-generation models for video generation and 3D object creation.
Legal Challenges and Ethical Debates
Midjourney finds itself at the center of growing legal and ethical controversies:
- Multiple ongoing lawsuits allege the company trained its models on billions of copyrighted images without permission
- Artists argue this constitutes large-scale intellectual property theft
- The outcomes could set important precedents for AI copyright law
In response, Midjourney has begun implementing optional filters to prevent direct style mimicry of living artists, though critics argue these measures don’t go far enough. The company maintains that its technology represents a new form of transformative creation rather than simple replication.
The Future of AI Art Generation
With V7, Midjourney reaffirms its position as an industry leader in AI-powered creativity. The update delivers tangible improvements that both casual users and professional artists will appreciate:
- Greater creative control through personalization
- More reliable outputs for commercial projects
- Faster iteration with Draft Mode
As Midjourney expands into video and 3D generation, its technology could revolutionize fields like:
- Game asset creation
- Advertising and marketing
- Architectural visualization
- Film pre-production
The coming months will be crucial as Midjourney:
- Rolls out missing V7 features
- Navigates its legal challenges
- Potentially unveils its next-generation video tools
For now, V7 represents the most sophisticated AI image generator available to the public, pushing the boundaries of what’s possible in algorithmic art creation while raising important questions about the future of human creativity in the age of artificial intelligence.
