<h3 id="comfyui-workflow-(image-to-video)">ComfyUI Workflow (Image to Video)</h3><ul><li>required Custom Nodes:<ul><li><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_Eclipse">ComfyUI_Eclipse: Version 1.5.9, </a><a target="_blank" rel="ugc" href="https://registry.comfy.org/publishers/rvage/nodes/ComfyUI_Eclipse">ComfyReg</a></li><li><a target="_blank" rel="ugc" href="https://github.com/r-vage/ComfyUI_SmartLML">ComfyUI_SmartLML: Version 2.0.7</a>, <a target="_blank" rel="ugc" href="https://registry.comfy.org/publishers/rvage/nodes/ComfyUI_SmartLML">ComfyReg</a></li><li>Update rgthree (forked): <a target="_blank" rel="ugc" href="https://github.com/r-vage/rgthree-comfy">https://github.com/r-vage/rgthree-comfy</a><ul><li>fixed the fast bypasser and applied some of the community fixes</li></ul></li></ul></li><li>the v1 version of the workflow uses the Wan Video Wrapper custom node.</li><li>the others are native versions (comfyui nodes)</li><li>its able to extend the initial video, by calculating the entered loop value or the max frame_load cap</li></ul>

Wan 2.1/2.2 (STD | GGUF)

<ul><li>a workflow that can be used for i2v or t2v with a14b or wan2.1 14b models, i haven't tested with 5b but should work too</li><li>it allows to start the initial render with text to video and extend the video with i2v by using the last frame as new ref image. in that case the 2 additional model loader are needed.</li><li>2 presets are in the archive</li><li>added the create loop that i've found in the darksidewalker workflow, it's kind of working, in the end it looks like a morph but yea.</li></ul>

Files with SHA256: 95a4ae4d8e6ab6e3bc3976baef2b0ff27d3f577061f9b6070a0697771cf643cc