![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
I've been using Wan2.1 locally on my desktop for making image-to-videos (i2v) utilizing the Pinokio installer. I found its "one-click" installation/upgrade/management effortless and elegant, and it comes with numerous configurations for low vRAM systems - as low as 5GB via optimization. It's pretty freaking sweet. The videos turn out best when limited (currently) to 5-seconds, but since I can make as many as I want, I've been importing them into Premiere Pro, and capturing the last frame as the first frame for a subsequent video, ultimately stitching them together. I'm also using the logos I make in Stable Diffusion to animate them.
Pinokio also (now) comes with a built-in video upscaled, which is the preferred method over trying to brute-force create HD. One of their models, Fun InP, can create a 5-second video on the RTX4090 in ~90 seconds (full 480p takes me ~10-minutes), so there's lots of time for other fun stuff. In other AI-related news, HiDream (17B) is out, and the news on that looks fantastic, though I haven't had a chance to play with it yet myself. I also was approved for Envato's new i2v beta, and that's even more amazing because it renders a photo in realtime as you type so you can make changes on-the-fly before submitting; I've never seen anything quite like it.
