🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Been testing the model — best results come at 720×1280 and ~81 frames. But the inference time is long (even on an A100 it takes 18+ minutes, with fp16). Shorter clips work fine, but lowering the resolution below 480p drastically worsens quality. (A cute little pig just popped

Yes , AI and CGI can work together ! Not against ! I made all this using ComfyUI with Wan 2.1 14B model on a H100. So the original 3D animation was made for game of thrones (not by me), and I transformed it using multiple guides in ComfyUI. I wanted to show that we can

Congrats!

Introducing A Branding New ComfyUI Look and Native API Nodes! Seamlessly integrate 11 amazing online visual AI models into ComfyUI, including Black Forest Labs Flux, Hailuo AI (MiniMax), MiniMax (official), Kling AI, Luma AI, PixVerse, Recraft, Stability AI, Google Veo2, Ideogram,

Kohya Tech It’s not a direction we are going towards, just an additional feature. We have been asked by a lot of users to allow them to use close sourced SOTA models. It’s more of a feature to allow comfy to do anything. Huge fan of your scripts 🙏🙏🙏

This is just the beginning! Happy to help contribute a lot of content and communications for this feature! API nodes are so Mac friendly lol

Hakoniwa is always the best! 🙌

🚨I made a ComfyUI Wrapper for NormalCrafter (Video normals) It's just the one node, extremely minimal reqs.txt to install and it auto downloads all the models for you on the first run. Please give it a try and let me know if it works for you LINK IN FIRST COMMENT BELOW 👇

Finally the powerful official version of Wan VACE! The team is working on more workflows and will share through the doc. An extra surprise is generating music via ACE Studio Ace Step. Super fast and allows detailed description about the song. The music below was generated in

Tried Veo3 to generate a scene in Hollow Knight and the audio effect is crazy! Please, we would like image ref input. Would def play with the prompts a bit!

When ComfyUI API node?

I am fully shocked

I believe virtual imagery allows for more creative freedom than live-action. As Satoshi Kon said, “In animation, only what is intended to be communicated is there.” Fictional visuals offer a greater potential to convey the creator’s intent — especially when you have strong

JO. Z Your comment questions the essential nature of creativity. In current creative activities, humans taking the lead aligns with the current approach to creativity where we 'materialize the images in our own minds.' I can agree with this. However, in the future, different forms of

Meanwhile in the open-source end of the pool, here's Wan + VACE + a style frame in ComfyUI 🤘 #aivideo Process images in thread 🧵

I really like how I can partially execute the node graph now, and then go down one rabbit hole deeper and deeper. Flux.1 Kontext is a really powerful model. The style is good. It follows when you tell it to keep the composition. I am chaining the local flux and online ones back

⚡️ 4-Hour ComfyUI Mini Hackathon Next Week!⚡️ 🚀 Drop in, build fast, win big. 🏆 Prizes: 2 brand-new RTX 5090 GPUs!! 🧩 Challenge: Ship product that uses ComfyUI, Vibe code a custom node, or Craft the slickest workflow content 🎉 After-party guaranteed ⏳ Spots are limited