ComfyUI

YUI is the first animated short film made with Comfy Agent. We partnered with Tokyo AI animiation artist 852話 Hakoniwa in this experiment. She made the full 4:20 film in 3 focused days of work. And we shaped the character together. YUI is the last three letters of ComfyUI. The behind-the-scenes video, also generated with Comfy Agent, was made to echo the theme "The story does not end". Comfy Agent handled shot regeneration, side-by-side model comparisons, and character consistency checks. That used to be weeks of manual testing and generation. When you are the main character, what will you do if your story ends in 3 minutes? Detailed behind-the-scenes article about our story with the artist, the creation ideas, and the production process: https://lnkd.in/eymcFrrb

Comfy Developer Platform: Building with Comfy API & Comfy Router Comfy Developer Platform is here! We'll give a quick overview of the features that make up Comfy Developer Platform, and then we're opening it up to YOU our community to build as a part of Open Call, our Comfy Dev Platform Challenge including a special night for building live in San Francisco on October 13th. Challenge details here: https://lnkd.in/gg9RE57Y Register for the 10/13 SF event here: https://luma.com/5mydmvu6 What we'll cover: 🤖 Intro to Comfy Developer Platform: Comfy API & Comfy Router 🧠 Open Call 🤖 Community Q&A — bring your questions ⏰ Subscribe and enable notifications to catch the stream live 💬 Submit your questions in chat during the broadcast or before the stream in the comments! 👍 Leave a like if you'd like to see more real-world agent workflows! https://comfy.org/platform

Comfy Agent is now available to everyone in Comfy Cloud. It plans, builds, and fixes ComfyUI workflows right on your canvas. We strongly believe creators shouldn’t choose between ambition and complexity. Comfy Agent is our answer. While it handles the technical part of Comfy, you can finally focus on visual craft! - It builds while you edit. It sees the assets you see. - Ask anything. Reference anything. - Create your own skills, or use the public ones. Agent in Comfy Desktop is coming soon in a few weeks. Click the link below to try Comfy Agent today ⬇️

One useful way to approach complex AI video shots is to separate **scene control from visual appearance**. For this 13-second action sequence, a Blender blockout establishes the underlying structure of the shot: camera movement, framing, character placement, movement trajectories, choreography, and timing. That blockout then becomes the driving reference in ComfyUI. Instead of asking the model to determine where six attackers should appear or how the camera should move, those decisions are already represented visually. Separate image references can focus on what the hero, enemies, environment, and final anime style should look like. This creates a useful division of responsibilities: **Blockout → motion, timing, camera, and spatial relationships** **Reference images → character identity, clothing, environment, and style** For shots involving multiple characters and coordinated action, blocking the scene first can provide much more explicit visual guidance than trying to describe every spatial relationship through prompting alone. Want to try this workflow yourself? Link in the comments below. https://lnkd.in/eMBbzbDi

Ideogram 4.5 is now available in ComfyUI. Anyone who has edited the same image five times in a row knows the problem. Colors shift, textures soften, and small artifacts pile up until the image is unusable. Ideogram 4.5 is built to stop that drift and preserve detail across multi-turn edits. What stands out: → Edit at full resolution. Work on a crop of a 24MP image and stitch it back in, so the rest of the frame stays sharp for close-ups and large-format print → Rewrite or translate text inside an image while keeping the original design → Restyle interiors, refresh product shots, and swap colorways without regenerating the whole scene → Use style references, sketch-to-image, depth-to-image, and old-photo restoration In ComfyUI, it slots right into your pipelines. You can chain edits, batch colorway variations for a campaign, or pair it with upscaling for print. Click the link below to run the workflows ⬇️

Deploy your ComfyUI workflow in minutes. If your team has to crawl through glass to run ComfyUI, you should try Comfy API.

Comfy API is live Until now, shipping a ComfyUI workflow to production meant rebuilding its environment somewhere else. Renting GPUs, reinstalling every custom node and model, untangling Python dependencies, and writing your own scaling logic. Comfy API handles that for you: → Builder reads your workflow JSON, finds the models and custom nodes it needs, and helps resolve dependency conflicts → Each Build pins the ComfyUI version, nodes, models, and Python deps together → Releases are immutable, so the environment you test is the one you deploy → Endpoints autoscale, down to zero when idle or with warm workers when latency matters → GPU time is billed by the second on RTX PRO 6000, H100, H200, or B200 Your graph stays yours, the engine stays open source, and your Builds stay portable. Click the link below to deploy your first workflow today

ComfyUI for Animation: Keyframe Exploration One of the slowest parts of pre-production is figuring out how a character reads from different angles before anyone commits to a shot. This workflow speeds that up. Start with a single character reference, choose any camera angle, and get a finished keyframe illustration back. How it works: → Load your character or character sheet into the input node. → Set the shot with the Multi-Angle Camera node by dragging the direction, elevation, and distance controls. The node writes the camera description for you and passes it to GPT Image 2 as the prompt. → Optionally, add a background to place the character in a scene, and describe a pose, expression, or action in the Additional Action node. You can switch off either one. → Run the workflow. GPT Image 2 combines the character, angle, and scene into a single on-model keyframe. For animators, storyboard artists, and studios, this makes it easy to try a range of shots quickly and stay consistent with the character design before moving into motion. Click the link below to run the workflow ⬇️

Recraft V4.1 Flash is now live in ComfyUI. It's the fastest image model on the market. It goes from request to finished image in 1.3 seconds, and it costs about 5× less than standard models. Try a direction, react to it, and try again. Do it dozens of times before a slower model would finish a handful. Recraft calibrates its models with design professionals, and Flash keeps that quality across photography, portraits, mood, complex scenes and early concepts. When a generation is worth keeping, Refine turns that exact image into a Pro-quality version at 2048×2048. It's a real detail pass, not an upscale. Run the model today with the link below ⬇️