- 版本 [version]
- 下载 131
- 文件大小 23446
- 文件计数 1
- 创建日期 2026年8月19日
- 最后更新 2026年9月21日
Phantom_FusionX_LoRA
Phantom_FusionX_LoRA 是一套面向视频生成场景的 ComfyUI 工作流,全图共 24 个节点、25 条连线。链路以 LoadImage ×4、ImageBatch ×3、CLIPTextEncode ×2 为主干,输出视频,适合需要稳定批量出图的同学。主模型 wanvaceWan2.1_T2V_14B_FusionX_VACE-Q3_K_S.gguf、wan2.1umt5_xxl_fp8_e4m3fn_scaled.safetensors、Wan_2.1wan_2.1_vae.safetensors。已挂载 LoRA:Wan2.1Phantom_Wan_14B_FusionX_LoRA.safetensors。下载后直接导入即可使用,建议显存 16GB 起(视频生成建议 24GB)。
- 主模型:wanvaceWan2.1_T2V_14B_FusionX_VACE-Q3_K_S.gguf、wan2.1umt5_xxl_fp8_e4m3fn_scaled.safetensors、Wan_2.1wan_2.1_vae.safetensors
- LoRA / 适配器:Wan2.1Phantom_Wan_14B_FusionX_LoRA.safetensors
- 核心节点:LoadImage ×4、ImageBatch ×3、CLIPTextEncode ×2
- 输出形式:输出视频
- 规模:24 节点 / 25 连线
- 建议显存:16GB 起(视频生成建议 24GB)
- 适用场景:图生视频、首尾帧、数字人与长视频一致性控制工作流。
工作流节点图
图较大可横向滚动;点击图片查看高清矢量原图(浏览器中可自由缩放)。
资源头图

使用说明
下载工作流文件后在 ComfyUI 中直接拖入画布(或菜单 Load 导入)即可运行;缺失的自定义节点用 ComfyUI Manager 一键安装,模型按文件名放入对应的 models 目录。
源码预览
{
"id": "9a6bb580-900f-4a5f-9233-12eb0fe41226",
"revision": 0,
"last_node_id": 319,
"last_link_id": 162,
"nodes": [
{
"id": 297,
"type": "UnetLoaderGGUFDisTorchMultiGPU",
"pos": [
-120.27359771728516,
-1230.2701416015625
],
"size": [
413.7545166015625,
156.38363647460938
],
"flags": [],
"order": 0,
"mode": 4,
"inputs": [
{
"label": "unet_name",
"localized_name": "unet_name",
"name": "unet_name",
"type": "COMBO",
"widget": {
"name": "unet_name"
},
"link": null
},
{
"label": "device",
"localized_name": "device",
"name": "device",
"shape": 7,
"type": "COMBO",
"widget": {
"name": "device"
},
"link": null
},
{
"label": "virtual_vram_gb",
"localized_name": "virtual_vram_gb",
"name": "virtual_vram_gb",
"shape": 7,
"type": "FLOAT",
"widget": {
"name": "virtual_vram_gb"
},
"link": null
},
{
"label": "use_other_vram",
"localized_name": "use_other_vram",
"name": "use_other_vram",
"shape": 7,
"type": "BOOLEAN",
"widget": {
"name": "use_other_vram"
},
"link": null
},
{
"label": "expert_mode_allocations",
"localized_name": "expert_mode_allocations",
"name": "expert_mode_allocations",
"shape": 7,
"type": "STRING",
"widget": {
"name": "expert_mode_allocations"
},
"link": null
}
],
"outputs": [
{
"label": "MODEL",
"localized_name": "MODEL",
"name": "MODEL",
"type": "MODEL",
"links": []
}
],
"properties": {
"cnr_id": "comfyui-multigpu",
"ver": "1.7.3",
"Node name for S&R": "UnetLoaderGGUFDisTorchMultiGPU",
"widget_ue_connectable": []
},
"widgets_values": [
"wanvaceWan2.1_T2V_14B_FusionX_VACE-Q3_K_S.gguf",
"cuda:0",
0.1,
false,
""
],
"color": "#322",
"bgcolor": "#533"
},
{
"id": 298,
"type": "MarkdownNote",
"pos": [
333.2388610839844,
-1248.0262451171875
],
"size": [
1014.274658203125,
420.11279296875
],
"flags": [],
"order": 1,
"mode": 0,
"inputs": [],
"outputs": [],
"title": "Low vram",
"properties": {
"widget_ue_connectable": []
},
"widgets_values":
… (完整内容请下载文件查看)
代码较长,以上仅前 3800 字符 —— 点击下方按钮下载完整 JSON。



