Higgsfield Genjutsu can replace people, products or scenes in a video based on reference images, charging 104 credits for 15 seconds at 720p, and supports 4-30 second source videos and up to 30 reference images.
Higgsfield offers two no-code ways to automate AI video creation: batch-generating videos in ChatGPT via the Higgsfield Extension and in Claude via Higgsfield MCP, or running recurring tasks on a schedule with Supercomputer; both consume regular plan credits and require no API key.
oh-story-dsh is a community plugin for DeepSeek Harness; once installed, DSH Web gains four workbenches for novels, short dramas, games and video commentary, and created content exists as real local files in the workspace rather than being written to a parallel database.
This guide compares five one-stop AI video platforms—Higgsfield, Runway, Kling AI, Dreamina CapCut, and LTX Studio—through a ten-step workflow, based on each company's public plan and documentation pages.
T8 has open-sourced a growth-level master SKILL that can be paired with MiniMax H3 and any Agent to replicate viral videos; the company says it works better the more it is used. The SKILL comes with a T8 cloud image and ComfyUI integration package V10, and lists several trainer images, including the Wan2.2 DP all-purpose trainer and the Qwen Image Ai toolkit full-featured trainer.
歸藏 says Anatomy is the hottest Skill this week, able to turn complex mechanical structures into distinctive isometric 3D animations. Combining images and videos generated with the Grok bot and Jimeng, he used the Skill to make a science video explaining how chips are manufactured. The cited Anatomy introduction shows it can generate rotatable, inspectable isometric wireframes, supports WebGL shaders such as fire, water, and plasma plus real 3D motion, and can be installed as an agent skill and is fully open source.
Higgsfield released Katana, an editing skill that runs inside Claude, calling Higgsfield models through the Higgsfield MCP connector; using a reference video or a preset from the preset library plus up to 5 photos of people, products, or clothing, it reconstructs a new video that preserves the reference clip's shot order, duration, and audio track, with results appearing in the conversation and in Higgsfield Assets.
Why it matters: Hand the reference clip's shot order, rhythm, and audio-track structure to Katana inside Claude to replicate it, so creators can judge whether manual matching cuts can be skipped.
fal's minimax/h3-max/relight can now be called directly in ChatGPT and Codex, letting users generate media without leaving the conversation. ComfyUI-MiniMaxH3-Studio supports turning descriptions into MiniMax H3 videos within ComfyUI, with a built-in editor for editing and rendering. In addition, PlateExtract, fine-tuned on Qwen Image 2.1, can extract transparent PNG foregrounds from composite images.
Jue.com summarizes six pieces of content from October 12, five of which are open-source tools that run locally: the short-drama generation canvas Toonflow, the visual scraper EasySpider, the Windows mouse wheel StarPie, the Bilibili concurrent stream-fetching Thread Ripper, and CodexHost, which packs multiple Agent Harnesses into Codex Desktop; the other piece covers unlocking Gemini membership features.
MiniMax H3 is a video generation AI model released by MiniMax in July 2026; it can generate video and audio simultaneously (including Japanese dialogue) and supports Turbo LoRA to reduce generation steps. In ComfyUI 0.30.0 and above, using the official template, you can generate a short video from a single image by simply placing the image (first frame) and describing the motion with a prompt, making it suitable for complete beginners.
The author compared 44 video upscaling and restoration methods on ComfyUI, all running locally for free, and distinguished between two scenarios: restoration and upscaling. Restoration targets low-quality footage from small cameras, surveillance, and the like, while upscaling targets upgrading HD to FHD, 4K, and 8K and recovering skin texture and eyelash detail. The article releases a large-scale comparison workflow that can run all 44 methods in one go, making it easy to pick suitable methods for different content such as people, landscapes, or animation.
Why it matters: The author separates the two needs of restoration and upscaling, uses the same footage to compare 44 methods side by side, and releases a ComfyUI workflow that can run in one click.
The author tested MiniMax H3's RefMod custom node in ComfyUI. It compresses character images, motion videos, or audio into reusable reference files, requires no LoRA training, is used much like a LoRA loader, allows adjusting strength, and can combine multiple references to put characters in the same frame.
Why it matters: The author tested MiniMax H3's RefMod custom node and provides the complete parameters and prompt-writing approach, from installation and building a reference library to integrating with the official T2V workflow.
Simple Scribble Canvas makes its first release, providing ComfyUI with centered canvas zooming, per-color brush size settings, and embedded workflow drawing.
AI Canvas v0.10.6 optimizes media memory and canvas recovery in performance mode, tightens idle media caching and preloading scope, and reduces continuous drawing and dynamic effects to lower the load on low-spec devices. Generated files with the same name are suffixed _1, _2 in sequence; a fix addresses failure to open the standalone assistant window; and prompt editing adds separate temporary undo and redo.
JJYB_AI_VideoAutoCut v3.3.1 adds a bilingual Chinese-English interface (i18n); 200+ interface strings across the site support one-click switching between Simplified Chinese and English, the language setting is persisted, and the default Chinese display matches the original version.
ScreenWeaver launched an MCP server, allowing Claude, ChatGPT, Claude Code, Cursor, and VS Code to directly read and write scripts and storyboards in projects via OAuth, with no copy-pasting needed.
When the author ran MiniMax H3 video generation with ComfyUI on an A40 on RunPod, 12 consecutive 10-second videos worked fine, but after switching to 12 seconds, the second video would invariably crash within tens of seconds to a few minutes, with the only error being Connection refused.
A creator used ChatGPT to generate about 25 storyboard images, free voiceover from Irodori-TTS, a rented GPU on Runpod to run miniMax H3 for video generation, and capcut for editing to make a one-minute AI animation, at a cost of about ¥90–¥150 and taking two days.
The author runs MiniMax H3 on an A100 in Google Colab and calls it from local Codex; in actual testing, generating a 960×544 10-second video took about 6 minutes 32 seconds.
DramaClip releases v0.2.1, adding a visual track that uses local Qwen3-VL models in three tiers (2B/4B/8B, GGUF quantized + llama.cpp, runnable on CPU) to uniformly sample 16 frames per episode and generate timestamped visual evidence for script prompts to reference by episode.
Using Opus 5.5 combined with video models to attempt narrative film expression, all arrangement, video effects, prompts, and editing were decided by Opus 5.5, while Jimeng CLI handled image and video generation for the corresponding storyboards. A large number of code-based animation effects were also overlaid on the video, noticeably improving the texture, and the author says the result is good.
ComfyUI creative project bkwildcards has released an update, adding the new Amano Style dark ink-wash watercolor style; the total number of art styles now reaches 13, and they appear simultaneously in both Art Style dropdown menus.
This tutorial explains how to set up an environment on Windows using the ComfyUI integrated package from Bilibili creator 秋葉aaaki; after unzipping, double-click the one-click launcher to open the interface. Built-in templates can be used directly, but models must be downloaded separately. The author uses WorkBuddy to import the integrated package folder and completes the initial configuration through conversation, downloading the qwen-image-2.1-turbo workflow and setting up text-to-image, image-to-image, and image-to-video, among others; WorkBuddy can also automatically build and execute workflows via API.
Aimed at the need to bring AI image generation in-house into one's own production environment, the article outlines the division of labor between the two routes of Stable Diffusion and FLUX: the Stable Diffusion family is suited to running on local or cloud GPUs.
AITuber releases a guide to choosing AI content repurposing tools, matching tools to source material: for articles or PDFs, choose AITuber to generate narrated videos with visuals and subtitles; for long videos, choose Opus Clip; for podcasts, choose Castmagic; for distributing finished clips, choose Repurpose.io.
The ArchiGen article argues that when a project's relationship with the city is itself a design proposition, the urban context should be built as a navigable model rather than a static backdrop. It cites a Chaos-sponsored Architect Magazine case: Morevi Studio's 84-hectare proposal for Ambassadori Island Batumi in Georgia builds Batumi and the artificial island as two linked zones placed in the same scene, enabling real-time review of cameras and atmosphere.
Chaos integrates Veras into V-Ray, supporting SketchUp, Rhino and 3ds Max; users can send viewport screenshots or VFB images to Veras and use Nano Banana 2 to generate style options, without modifying the 3D scene or geometry.
OpenAI Codex can use CoAnimator to edit video projects and call the renderer to output MP4: Codex is responsible for reading requirements, planning scenes, writing project files, and making revisions based on feedback, while CoAnimator handles preview and rendering.
Wan 2.2 image-to-video can run on an RTX 5070 with 12GB of VRAM; a 57-frame clip takes 10 minutes 38 seconds at 832×480 and about 67 minutes at 1536×864, both including frame interpolation and upscaling to 2048×1152.
The author used Opus 5.5 with three.js to basically produce a rainy-night convenience store miniature scene in one round, with only clipping fixed in the second round, all within the 5 hours limit of the cheapest membership. Prompt highlights include a third-person view miniature 3D model that can be dragged to rotate and zoom, a Japanese-style convenience store street corner as the main subject, a warm-cool contrast between the bright warm interior and the wet rainy night outside, cel-shaded cartoon rendering, and animated touches such as continuous rainfall and puddle ripples.
Fable 5.5, in gray release, works almost entirely code-driven in Claude Code, producing a poetic animation in 2.5 hours, with art direction, style, pacing and storyboard all decided by the model itself. The work is based on imagery from Laufey's "Promise" and "5 Centimeters per Second," featuring a hand-drawn brushstroke opening, subtitles in Chinese, English and Japanese, and a dot-matrix map shot of a 7.18m walk; only some images and music are external assets.
The author trained a Qwen Image 2.1 character LoRA with ai-toolkit 0.13.19 on Google Colab's A100 80GB; 1,000 steps took about 21 minutes this run, whereas the same number of steps on a local RTX 4070 12GB takes about 2 hours.
Stable Diffusion Art publishes a tutorial on how to build a MiniMax H3 text-to-video workflow in ComfyUI, generating a 5-second 24fps clip with stereo sound in one pass.
Aitrepreneur releases a tutorial demonstrating how to train a LoRA for Qwen Image 2.1 EDIT with AI TOOLKIT locally or on RunPod, using under 10GB of VRAM. The video also covers building your own training dataset, setting training parameters, and comparing the trained LoRA's results in ComfyUI.
The ComfyUI creative project Gallery Loader releases an update: the picker now supports audio card previews with a play button (#124). On security, absolute path reading is restricted to ComfyUI's own directory (#127), and a bug is fixed where a full list operation in xmp would clear its own rating cache (#123).
Lanshu open-sourced Agent Skill lanshu-create-ai-presenter-video; in Claude Code, Codex, Gemini CLI, Cursor, and OpenCode, a single sentence runs the entire process of script, voiceover, digital human generation, lip-sync, subtitles, editing, rendering, and quality check. The repository has 2.6k Stars, 404 Forks, and an MIT license.
Why it matters: An Agent Skill that breaks AI video production into computable states and paid gates; its workflow design approach can be transferred to other content pipelines.
AI filmmaker Frank Houbre disclosed his 2026 creative tool stack, divided into six stages according to the filmmaking process, with only one main tool kept for each key stage. The writing stage centers on ScreenWeaver, which he co-created, with Claude and Claude Code handling review and repetitive tasks; image and video generation center on Higgsfield as the main tool because it aggregates multiple models under the same account.