Spinoff of Yet another workflow — v1.0

Spinoff of Yet another workflow is for creating and preparing video from an incoming file with Wan Video 2.2 I2V-A14B; it can produce a video, first image, and extracted last image while supporting reusable prompts.

It extracts the last image for use when extending the video.

A prompt node stores reusable prompts and lets you combine prompt pairs with on/off switches.

The input is an incoming file; outputs include a video, first image, and extracted last image.

The incoming filename is copied to the video, first image, and extracted last image.

The video parts are put together and upscaled using Davinci Resolve.

The extra node needed for this workflow is described in the linked Civitai article.

Named node packs include ComfyLiterals, comfyui_fearnworksnodes, comfyui-art-venture, comfyui-custom-scripts, comfyui-easy-use, comfyui-kjnodes, comfyui-mxtoolkit, ComfyUI-SimpleSwitch, rgthree-comfy, tripleksampler, comfyui-wanresolutions, comfyui-videohelpersuite, RES4LYF, and

The listed model files include gimmvfi_r_arb_lpips_fp32.safetensors, lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank256_bf16.safetensors, umt5_xxl_fp8_e4m3fn_scaled.safetensors, and wan_2.1_vae.safetensors.

They also include wan2.2_i2v_high_noise_14B_fp16.safetensors, wan2.2_i2v_low_noise_14B_fp16.safetensors, wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors, and wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors.

Use the prompt node and on/off switches for reusable prompt pairs; the settings changed most often are grouped in a box at the top.

Suggestion · not verified

The described setup does not use T2V and uses the Clownshark sampler; verify both settings for your use.

Suggestion · not verified

Before use, verify which package provides ComfyUI-GIMM-VFI, film-grain-ltxv, and tripleksampler.

Suggestion · not verified

Before use, verify the model source for gimmvfi_r_arb_lpips_fp32.safetensors, lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank256_bf16.safetensors, wan_2.1_vae.safetensors, wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors, and wan2.2_t2v_lightx2v_4steps_lora_v1.1_low‌_‌‌

Suggestion · not verified