AI Video Mastery: ComfyUI & FramePack, Hunyuan, Wan2.2

Set up Hunyuan as a video checkpoint model to generate your first video content by using templates and workflows, and configure decoding nodes, text encoders, and diffusion models.

Learn to set up ComfyUI for Hunyuan t2v workflows by installing custom nodes, loading gjf models, using the video combine node, and configuring frame rate and tile settings.

Learn to create the first hunyuan video with comfyui using a pre-made workflow, monitor ram, vram, and gpu usage, and adjust output size for efficient anime-style generation.

Explore running and comparing ComfyUI videos with Hunyuan 2nd video, FramePack, and Wan2.2, using prompts, seeds, and three-second renders to assess quality, motion, and prompt adherence.

Build a custom ComfyUI node for video sizing and frame calculation using Quinn; configure aspect ratios, resolutions, fps, and optional image input via the AI Alchemist workflow.

Learn how Quen assists prompting in ComfyUI, crafting full positive prompts and managing chroma key green screen setups to create transparent video backgrounds for flexible frame pack workflows.

Develop a practical image resize workflow in ComfyUI using Convoy Essentials, compute the longest size with a max scale factor, and align outputs to multiples of 16 for video frames.

Create a frames counter in comfyui using a seconds float and an fps integer, then calculate total frames by multiplying and adding one, reusable as a node template.

Learn to build an image-to-video workflow in ComfyUI with Hunyuan I2V part 1, using custom nodes, image loading, prompts, and clip vision guidance.

Master ComfyUI image-to-video workflows by cropping and resizing with image crop and image resize nodes, integrating clip vision, convoy essentials, and auto-sized outputs for reliable previews.

Explore frame pack in comfyui by configuring forwards and backwards samplers, enabling time-stamped text prompts, and building a frame-by-frame workflow with downloaded frame pack and clip models.

Learn to build a FramePack B workflow in ComfyUI: load the 002 frame pack model, configure offload, attention, VA encoding/decoding with Hunyuan bf16, and export a video.

Run FramePack B in ComfyUI, tune the frame sampler, tcache, prompts, ambient interpolation, and resolution constraints to generate multi-frame videos efficiently while managing RAM usage.

Learn to configure comfyui frame pack t2v sampler, adjust frame rate to 24, remove warm-up frames, and use image or text-to-video prompts with dual clip loader.

Compare texture video approaches in frame pack, analyze 24-frame results, and optimize quality by adjusting tcache and sampler settings while evaluating F1 and B models.

Master time-based frame pack workflows in comfyui by building text-to-video sequences with f1 sampler, v loader, and frame pack timing, using empty images, seeds, and prompts for blended video content.

Create an image-to-video workflow in ComfyUI frame pack, loading and resizing an image, setting video size and length, and crafting prompts with negative prompts for transitions using the B1 model.

Master the comfyui framepack f1 i2v workflow by editing prompts, frame peg timing, and start image embeds to produce timed image-to-video sequences.

Explore the ComfyUI Wan 2.2 text-to-video workflow, loading high-noise and low-noise diffusion models (UMT XXL FP8, 12.1 VA), Laura's templates, and Unet loaders to produce a 16 fps MP4 video.

Learn how to visualize sigma curves in ComfyUI Wan 2.2, compare schedulers and beta samplers, and fine-tune high-noise to low-noise transitions for better animation quality.

Analyze ComfyUI Wan 2.2 T2V result of a dragon over a burning castle, focusing on beta sampler 57, steps, and cfg. Apply tips for animated previews, fixed inputs, and upscaling.

Explore using lightning lauras with 12.2 in ComfyUI Wan2.2, compare high noise and low noise models for text-to-video results, and test step and shift settings to balance movement and detail.

Adjust and compare ComfyUI Wan 2.2 t2v lightning results by tweaking steps, prompts, and schedulers to balance movement, color, and realism in text-to-video outputs.

Learn to build the image-to-video workflow in comfyui wan 2.2 I2V, selecting high/low noise checkpoints and configuring diffusion models, frame rate, and prompts.

Explore hands-on testing of image-to-video workflows in ComfyUI Wan2.2, using bypass, shift, and middle-step automation with integer math, plus comparing samplers and schedulers for optimal results.

Learn to extend videos in ComfyUI by loading last frames or videos, using image-to-video workflows, and batching frames with color matching to create longer, coherent outputs.

Compare the outcomes of ComfyUI and FramePack alongside hunyuan across prompts, noting movement, quality, and coherence. Evaluate time-based framing and B model performance for long videos and text-to-video results.

Upscale videos in ComfyUI by loading the video and selecting an upscale model (2x or 4x). Install models, align frame rate with video info, and export MP4 with CRF compression.