Today we are launching VE2 — a brand new model for Visual Electric. It can produce hyper realistic photos, accurate text, and generate images 2.5x faster. This is the biggest update ever to Visual Electric. 🔊 SOUND ON

One of the biggest limitations of our previous model was the ability to produce photorealistic images with accurate human anatomy and realistic textures. With VE2, it's becoming difficult to recognize AI generated photos. (Prompts are in the ALT)

VE2 has much better prompt adherence and can images with typography and lettering. For example, "A woman in a red blouse and blue jeans, on a street corner, holding a newspaper."

VE2 can also generate images with much more accurate typography and lettering.

We are also introducing fast mode, which will generate images 2.5× faster. It's great for exploring ideas quickly in lower fidelity. I've found that in some cases I actually prefer the aesthetic of the fast mode model.

And finally, we are also launching 26 new and updated styles that make use of the model. Over the coming weeks, we will be transitioning all of our styles to the new model. Look for the VE2 badge in the top right corner.

You can learn more about VE2 on our Product Hunt page.

is this a model you guys trained from scratch?

So excited to try this. Realism is what’s been limiting for me so far in VE. Congrats on the launch!!

Can’t wait to hear your thoughts ⚡️

@jsngr WTF this is so good