According to Midjourney’s official subreddit, V5 includes new features such as:

• Much wider stylistic range and more responsive to prompting
• Much higher image quality (2x resolution increase) improved dynamic range
• More detailed images with details more likely to be correct and less unwanted text.
• Improved performance with image prompting
• Experimental features such as seamless tiling and custom aspect ratios

They go on to explain:

“V5 is our second model trained on our AI supercluster and has been in the works for 5 months. It uses significantly different neural architectures and new aesthetic techniques. V5 isn’t the final step, but we hope you all feel the progression of something deep and unfathomable in the power of our collective human imagination.”

https://www.midjourney.com/
https://www.reddit.com/r/midjourney/comments/11s7ezp/welcome_to_the_new_v5/
https://twitter.com/midjourney/status/1636130389365497857