Midjourney’s new version + SVG graphics generation

Midjourney V4 is out now.
I’ve been a subscriber to Midjourney ever since their original beta and watching the progress has been pretty incredible. If you haven’t tried the new version, you can do so right now in the Midjourney Discord.
To use the new V4 model with your generations, simply add the –v 4
flag to your prompts. Here are some results I got last night:


SVG graphics generation
A friend of mine, Ajay Jain, recently extended diffusion models to also work with SVG graphics (e.g. vector-based images.) These allow for infinite resolution and are used extensively in the graphic design industry. Check it out here:
It's alive! Since there's interest, I released my code for automatically generating SVG vector graphics from textual captions: https://t.co/U39Ak5FiO1. Based on @tzumaoli's diffvg and @OpenAI's CLIP. Caption: "a painting of an evergreen tree" pic.twitter.com/3c6DBMdwFC
— Ajay Jain (@ajayj_) March 9, 2021
On a side note, I am only 10 Twitter subscribers away from 2,000. It’s kind of crazy to see how much my account has grown since I just started posting things I was interested in.
Ok, that’s all for now! Talk to you again soon.
-John