【WAN2.1】TXT to VIDEO — v2.1

This ComfyUI workflow is for generating video or images from a written prompt and a loaded Wan Video model; it can optionally upscale images and save images with generation parameters in their metadata.

The v2.1 download includes three workflow files—“WAN2.1 - TXT to VIDEO (base).json,” “WAN2.1 - TXT to VIDEO (gguf).json,” and “WAN2.1 - TXT to VIDEO 2.1 (FLUX+gguf).json”; the base and GGUF files each list 55 nodes, while the FLUX+GGUF file lists 83.

Enter a written prompt and load a model; the workflow also accepts LoRA modules through three slots.

It can produce text-to-video or text-to-image results, optionally upscale images with the built-in SeedVR2 upscaler, and save images with generation parameters in their metadata.

Enable “Use Nodes V2 (Vue)” in ComfyUI Settings, refresh the page, and then load the workflow; the listing warns that leaving it off may cause display problems.

Suggestion · not verified

Install ComfyUI-UmeAiRT-Toolkit through ComfyUI Manager by searching “UmeAiRT,” or use the UmeAiRT Auto-Installer.

The workflow entries list these ComfyUI packages: comfyui-custom-scripts, comfyui-easy-use, comfyui-frame-interpolation, comfyui-impact-pack, comfyui-kjnodes, comfyui-multigpu, comfyui-mxtoolkit, ComfyUI-Upscaler-Tensorrt, comfyui-videohelpersuite, rgthree-comfy, and was-node-s e

The FLUX+GGUF workflow additionally lists comfyui-florence2.

The listed model files vary by workflow: base—rife47.pth, umt5_xxl_fp8_e4m3fn_scaled.safetensors, wan_2.1_vae.safetensors, wan2.1_i2v_480p_14B_bf16.safetensors; GGUF—rife47.pth, umt5_xxl_fp8_e4m3fn_scaled.safetensors, wan_2.1_vae.safetensors, wan2.1-t2v-14b-Q5_K_M.gguf; FLUX+GGUF

For manual setup, place base-version fp16 or fp8 T2V models, or GGUF-version Q8, Q5, or Q3 T2V quant models, in models/diffusion_models. Place umt5_xxl_fp8_e4m3fn_scaled.safetensors in models/clip, wan_2.1_vae.safetensors in models/vae, lightx2v_T2V_14B_cfg_step_distill_v2_lora_…

You can turn SeedVR2 on or off. The three LoRA slots each have on/off and strength controls, and additional LoRA modules can be connected.

Before downloading or running a variant, verify that the page shows the title “【WAN2.1】TXT to VIDEO — v2.1” and publisher “UmeAiRT,” and that the model files and ComfyUI packages for that variant are installed and matched.

Suggestion · not verified

No license is specified in the listing, so check applicable terms before reuse.

Suggestion · not verified