Skip to content
Date and source

Dates follow the news timeline in Beijing time and include both endpoints.

Oct 19Mon
  1. Reddit comfyui · 创作与工作流分享17

    Utsushi: a Higgsfield clone project, rewriting Comfy Router and Comfy SDK in Go

    While participating in the open-source Higgsfield competition, developer _playlogic_ rewrote Comfy Router and Comfy SDK in Go, building on a self-developed GO/HTMX open-source framework without using React or Next.js. The tool can connect to any version of Comfy and supports Cloud, Runpod, local servers, and desktop, and is still in development with new features being added.

  2. Reddit comfyui · 创作与工作流分享38

    tscomfy: write ComfyUI workflows as code in TypeScript, with types generated from your own server

    The developer released the open-source TypeScript library tscomfy, which reads the ComfyUI server's /object_info to generate types for installed nodes (including custom node packs); workflows are written as code, and wiring errors, misspelled checkpoints, or missing inputs are reported in the editor, with autocompletion for models and samplers.

  3. Reddit StableDiffusion · 创作者与工作流分享60

    ComfyUI-MOSS-Prompt-Enhancer: brings the MOSS-Image prompt enhancement pipeline into ComfyUI

    The author released MOSS Prompt Enhancer, a custom node that brings the MOSS-Image prompt enhancement workflow into ComfyUI. In structured layout mode, it expands a simple prompt into a description, then converts it into structured JSON, which is validated and compiled into a MOSS-usable prompt; a local GGUF model supplements subject, scene, lighting, and composition details, and generates a layout with descriptions and bounding boxes.

  4. ComfyUI Wiki · 工作流与开源创作分享65

    rapid-moss 4-step LoRA frees MOSS-Image-SFT from the 50-step sampling loop

    darask0 released the rapid-moss 4-step adapter on October 17, 2026, freeing MOSS-Image-SFT from the 50-step, CFG 3 sampling loop, along with a ComfyUI custom node that includes its own sampler.

    Why it matters: Two LoRAs compress MOSS-Image-SFT's 50-step sampling to 4 steps; the article provides node wiring and VRAM usage to help evaluate replacing existing image generation pipelines.

  5. LichtFeld-Studio 官方发布22

    LichtFeld Studio 0.5.4.1 released: bug-fix patch plus drag-and-drop export, training resume, and more

    LichtFeld Studio 0.5.4.1 is released, supporting dragging splats from the scene graph into folders or directly into Blender as PLY/SPZ, with the up axis selectable in preferences. This version adds training session resume, batch video frame extraction at the source frame rate, and numeric position, rotation, and scale fields for crop boxes and ellipsoids. It also fixes image loading when no GPU JPEG decoder is available, Python console freezes, and VRAM exhaustion when merging large compositions.

  6. Reddit StableDiffusion · 创作者与工作流分享30

    Running moss image on a 3050 4GB laptop: from 15 minutes optimized down to about 62 seconds

    A user ran moss image on a 3050 4GB laptop; the first image took about 15 minutes, which dropped to about 62 seconds after optimization (with prompt caching), and an experimental int8 option was added that supports switching without deleting the original model. New prompts take longer, and only a small number of images have been tested so far.

  7. ComeCut · 开源创作项目发布22

    ComeCut v0.2.0 releases its first Windows desktop version

    ComeCut v0.2.0 released its first desktop version, supporting Windows only, with a 30MB installer, fast startup, and low resource usage. This version adds a multilingual interface (English first, covering menus and the AI assistant), fixes the issue of audio being lost after separation and adds audio processing progress display, and also fixes known bugs. The development build is set to automatically expire after 30 days.

  8. CoAnimator · 可编辑视频与动画工作流20

    Roundup of the 7 best AI motion graphics tools in 2026

    CoAnimator released a guide to the 7 best AI motion graphics tools in 2026, ranked by use case rather than performance scores; the selected tools include CoAnimator, After Effects, Jitter, DaVinci Resolve, Spline, and Canva, with AI video generation tool Higgsfield in a separate section. The guide emphasizes the difference in control over later modifications between editable projects and generated clips, and recommends first excluding unsuitable tools based on delivery requirements, then testing creative features.

Oct 18Sun
  1. Reddit comfyui · 创作与工作流分享38

    Developer launches open-source Android app Comfy Workbench v0.3.0 to run ComfyUI workflows remotely from a phone

    The developer released the open-source Android client Comfy Workbench v0.3.0, which lets users import and run ComfyUI workflows on their phones. Users only need to start ComfyUI with --listen 0.0.0.0 and enter http://:8188 on the same Wi-Fi to connect, with no additional software required; remote access is optional.

  2. Stuubzzz · 视频工作流实测74

    Stubelius Ultimate H3 0.13 released: chunked generation of MiniMax H3 long videos in ComfyUI, with individual chunks redoable

    Stubelius Ultimate H3 0.13 is the author's open-source MiniMax H3 ComfyUI workflow, under the MIT license, made public on GitHub on 2026-10-18.

    Why it matters: The author released a ComfyUI workflow for chunked rendering and single-chunk redo, and provides measured timings on an RTX 5090, so you can judge whether it is worth switching your long-video production pipeline to it.

  3. NerdyRodent · YouTube 创作视频62

    Nerdy Rodent explains the workflow for using the open-source Ming Image design model in ComfyUI

    ComfyUI now supports the Ming Image Design model, offering a smaller INT8 version and a larger BF16 version. Ming-Image-0.1-Design is a 6B text-to-image model aimed at text-dense visual design such as UI, infographics and posters; it can generate complete images and supports RGBA output with transparent backgrounds.

  4. ai-film-skills · 开源创作项目发布30

    AI Storyboard Director 5.8.6 released: unified cinematic-quality and professional-optics version

    AI Storyboard Director 5.8.6 unifies the source code and the downloadable package; the six modules are in the order generation task, visual style, main subject, scene state, emotional goal, segmented script, with an unnumbered negative prompt appended at the end. This version synchronizes the prompt compilation contract for video production and short dramas, preserving existing plot, shot order, style and continuity, and does not fix three shots. Package structure, standalone operation, public documentation and multi-platform GitHub checks have been completed; download ai-storyboard-director.zip to get the complete storyboard skill.

  5. AI-Canvas-tauri · 公开创作工具发布32

    AI Canvas v0.10.10 released: integrates ToAPIs and one-click ComfyUI installation

    AI Canvas v0.10.10 adds connections to the ToAPIs China site and international site, which share an API Key and can be switched directly on the vendor card; the text catalog is limited to the specified 9 models. The Windows version supports installing the official ComfyUI portable version in a separate directory, including Python and runtime dependencies; after a public GitHub download times out, it switches to a China accelerator at most once. It also fixes video delivery recovery, model selection, and asset reference interactions.

  6. open-ai-canvas · 开源创作工具发布22

    open-ai-canvas v1.6.8.11 released: canvas Agent dynamic forms add multi_select

    open-ai-canvas released v1.6.8.11, adding multi_select to canvas Agent dynamic forms, supporting multi-select default values, stable option IDs, and array answers. This version also fixes the issue where video form fields were filtered out by the backend, so aspect ratio (e.g., 9:16/16:9) and resolution (e.g., 480P/720P) options can be selected and submitted normally. This upgrade has no database schema migration, but the form protocol change requires updating both Web and Backend.

  7. note · seisei_surumono · 图像视频创作分享15

    ComfyUI virtual human node development (3): improvements to character dialogue and LLM chat mode

    ComfyUI virtual human node development reaches its third installment, with the author continuing to improve the character dialogue system: adding new characters to expand conversations, adding an embedding atlas to visualize dialogue data, and adding a silent ordinary LLM chat mode. For voice, Gemma4 is now used to produce audible sound, with support for switching to whisper, though accuracy is still low; a text transcription prompt button was also added.

  8. note · dreamy_quail9288 · 图像视频创作分享22

    ChatGPT-driven PERSONAL ART MUSEUM: an AI art prompt generator that turns you into your own museum

    The author released PERSONAL ART MUSEUM, a ChatGPT-based prompt generator that reads past chat history and turns personal experiences and emotions into exclusive artworks, with titles and curator commentary. After 2 major revisions and 18 fine-tunings—20 updates in total—it was finalized as Ver.3.8, supporting three modes of play: single-piece creation, 10 art forms side by side, and horizontal long exhibition catalogs. The web version is free to use; just copy the prompt and paste it into your own ChatGPT to generate.

  9. Reddit StableDiffusion · 创作者与工作流分享38

    A less-restricted LoRA trained on Qwen-Image-2.1 that supports all-around poses

    A creator released a rank-64 LoRA fine-tuned on Qwen-Image-2.1 that can generate almost any pose, covering prompts and compositions the base model struggles with. The LoRA works for both text-to-image and image editing without a separate adapter; tested parameters are 25 steps, CFG 4.0, LoRA strength 1.0. It is still v0.1 with no formal benchmark yet, and the weights and inference code are public.

  10. Reddit StableDiffusion · 创作者与工作流分享22

    First time using ComfyUI to try the MOSS image model

    A beginner used ComfyUI with the MOSS-Image SFT BF16 model for the first time to generate an anime character composited into a realistic environment at 2560 × 1440 resolution on an RTX 5080 (16GB VRAM), with satisfying results. They got started with a .json workflow generated by ChatGPT, and issues such as distorted background text and abnormal character interactions remain.

  11. Reddit StableDiffusion · 创作者与工作流分享22

    Recreating a Rambo: First Blood Part II animated scene with Minimax H3 and ComfyUI

    A creator recreated an animated scene from Rambo: First Blood Part II using ComfyUI, Krea 2, and Minimax H3, with the prompt "2d 1980s saturday morning style cartoon". All shots took about 115 generations, completed overnight on a single computer without a data center, and were edited in ArtCraft EffectCraft.

  12. Reddit StableDiffusion · 创作者与工作流分享15

    MOSS Image generating 6 cats dancing: SFT 40 steps vs. Krea 10 steps comparison

    A user used MOSS Image with SFT, 40 steps, cfg3, euler/sgm parameters to generate an image of 6 cats dancing in pairs on a dance floor, with a courier-font paper label in front of each cat noting its breed. Under the same prompt, Krea generated a comparison image with the same settings, no LoRA, 10 steps, and ersde/beta.

  13. Reddit StableDiffusion · 创作者与工作流分享22

    Same-prompt comparison of anime scene generation with SDXL (WAI-Illustrious v17), Krea 2 turbo, and MOSS-Image-SFT

    A user compared anime scene generation across three models—SDXL (WAI-Illustrious v17), Krea 2 turbo, and MOSS-Image-SFT—at 1024×1024 using the same set of pure booru tag prompts (no style prefix, no quality words, no negative prompts), with 4 scenes × 2 seeds, all without LoRA.

  14. openscreen · 公开创作项目发布8

    OpenScreen v2.1.0-rc.5 released

    OpenScreen released v2.1.0-rc.5, fixing an issue where the first frame showed Full Camera ghosting when a clip's start time in the compositor was imprecise under f32. On the preview side, it fixed restarting the playback clock at the requested time on freeze frames, and recompositing paused frames at the display time.

  15. openscreen · 公开创作项目发布8

    OpenScreen v2.1.0-rc.6 released

    OpenScreen v2.1.0-rc.6 fixed an issue where rounding up the chunk start at cut points during export caused Full Camera trajectory residue, and added timeline tests asserting that trimmed clip boundaries equal the cut points under f32.

  16. openscreen · 公开创作项目发布28

    OpenScreen v2.1.0 released

    OpenScreen released v2.1.0, with GIF export about 30 times faster and with exact output. It adds an MCP checkpoint that can roll back an entire agent turn in one step, and MCP clients can now read and edit all projects rather than only the currently open project. Whisper transcription supports per-project custom names and terminology, and a Glass Lens glass cursor effect was added.

  17. openscreen · 公开创作项目发布8

    OpenScreen v2.1.1-rc.1 released

    OpenScreen released v2.1.1-rc.1, fixing an issue where the entire card was skewed in fill format in the compositor, and correcting the display of blur masks and motion blur on skewed fill cards. See v2.1.0...v2.1.1-rc.1 for the full changelog.