Model files for the shipped workflows

ComfyTV ships a curated set of ComfyUI workflows under workflows/<kind>/. Each one references one or more model files that ComfyUI itself must be able to find under your ComfyUI/models/ tree. This page lists, per workflow, exactly which files are needed and where they go.

If you’d rather plug in your own ComfyUI workflow (different model, different sampler, different LoRA stack), you don’t need any of these — see custom-workflows.md for how to drop in a custom workflow JSON + _preset.json.

ComfyUI normally auto-creates these subfolders on first run; if one is missing, create it manually. Folders below are relative to ComfyUI/models/.

Generate · Text

Workflow	Files	Folder
Local Qwen3 4B	`qwen_3_4b.safetensors`	`text_encoders/`

Download: https://huggingface.co/Comfy-Org/flux2-klein/tree/main/split_files/text_encoders

Generate · Image

Workflow	Files	Folder
Local SD1.5 / Local SD1.5 I2I	`v1-5-pruned-emaonly-fp16.safetensors`	`checkpoints/`
Image Ideogram4 T2I	`ideogram4_fp8_scaled.safetensors`, `ideogram4_unconditional_fp8_scaled.safetensors`, `qwen3vl_8b_fp8_scaled.safetensors`, `flux2-vae.safetensors`	`diffusion_models/`, `diffusion_models/`, `text_encoders/`, `vae/`

Generate · Video (LTX 2.3)

All four variants share the same base model + text encoder + LoRA stack.

Workflow	Files	Folder
Local LTX 2.3 T2V	`ltx-2.3-22b-dev-fp8.safetensors`, `ltx-2.3-spatial-upscaler-x2-1.1.safetensors`, `ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors`, `gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors`, `gemma_3_12B_it_fp4_mixed.safetensors`	`checkpoints/`, `latent_upscale_models/`, `loras/`, `loras/`, `text_encoders/`
Local LTX 2.3 I2V	same as T2V	same
Local LTX 2.3 FLF2V	`ltx-2.3-22b-distilled-fp8.safetensors`, `gemma_3_12B_it_fp4_mixed.safetensors`	`checkpoints/`, `text_encoders/`
Local LTX 2.3 IA2V	same as T2V/I2V	same

Download:

Generate · Audio

Workflow	Files	Folder
ACE-Step v1 Song	`ace_step_v1_3.5b.safetensors`	`checkpoints/`

ACE-Step v1 ships natively in ComfyUI core — no custom node needed.

Image edits

Workflow	Files	Folder
Flux Canny Edit (image-edit)	`flux1-canny-dev.safetensors`, `t5xxl_fp16.safetensors` (or `t5xxl_fp8_e4m3fn_scaled.safetensors`), `clip_l.safetensors`, `ae.safetensors`	`diffusion_models/`, `text_encoders/`, `text_encoders/`, `vae/`
Flux Fill Inpaint (inpaint)	`flux1-fill-dev.safetensors`, t5 + clip_l + ae (same as above)	`diffusion_models/`, `text_encoders/`, `text_encoders/`, `vae/`
Fooocus SDXL Inpaint (inpaint)	`juggernautXL_version6Rundiffusion.safetensors`, `fooocus_inpaint_head.pth`	`checkpoints/`, `inpaint/`
Flux Fill Outpaint (outpaint)	same as Flux Fill Inpaint	same
Fooocus SDXL Outpaint (outpaint)	same as Fooocus SDXL Inpaint	same
LaMa Erase (erase)	`big-lama.pt`	`inpaint/`
BiRefNet Cutout (cutout)	`birefnet.safetensors`	`background_removal/`
Ultrasharp 4x (upscale)	`4x-UltraSharp.pth`	`upscale_models/`

Download:

Flux dev: https://huggingface.co/Comfy-Org/flux1-dev
t5xxl / clip_l: https://huggingface.co/comfyanonymous/flux_text_encoders
VAE (ae.safetensors): https://huggingface.co/Comfy-Org/Lumina_Image_2.0_Repackaged
LaMa Erase: install Acly/comfyui-inpaint-nodes for the INPAINT_LoadInpaintModel node + big-lama.pt from its repo.
Fooocus inpaint head: bundled with the Fooocus inpaint patch (fooocus_inpaint_head.pth).
BiRefNet: native to ComfyUI core (LoadBackgroundRemovalModel + RemoveBackground); https://huggingface.co/Comfy-Org/BiRefNet

Image variations (single-workflow N-output)

The “multi-view” (Face / Product / Character / Multi-cam 9) and the “sequence” (Story 4 / Storyboard 25) presets share the same Qwen base — only the LoRA differs.

Group	Files	Folder
Multi-cam / 3-view set	`qwen_image_edit_2511_fp8mixed.safetensors`, `Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors`, `qwen-image-edit-2511-multiple-angles-lora.safetensors`, `qwen_2.5_vl_7b_fp8_scaled.safetensors`, `qwen_image_vae.safetensors`	`diffusion_models/`, `loras/`, `loras/`, `text_encoders/`, `vae/`
Story 4 / Storyboard 25	`qwen_image_edit_2509_bf16.safetensors`, `Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors`, `next-scene-qwen-image-lora-2509.safetensors`, `qwen_2.5_vl_7b_fp8_scaled.safetensors`, `qwen_image_vae.safetensors`	`diffusion_models/`, `loras/`, `loras/`, `text_encoders/`, `vae/`
Multiangle (3D-camera-driven)	same as the multi-view set above	same

Download:

Qwen Image Edit base: https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI
Qwen Image base: https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI
Lightning 4-step LoRA: https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning
Multiple-Angles LoRA: typically https://huggingface.co/PVC-Inc/Qwen-Image-Edit-2511-Multiple-Angles-LoRA
Next-Scene LoRA: https://huggingface.co/lovis93/next-scene-qwen-image-lora-2509

Relight

Workflow	Files	Folder
Flux2 Klein Relight	`flux-2-klein-9b-nvfp4.safetensors`, `Sun_direction_LoRA_Flux_2_Klein_9b_v1.safetensors`, `qwen_3_8b_fp8mixed.safetensors`, `flux2-vae.safetensors`	`diffusion_models/`, `loras/`, `text_encoders/`, `vae/`

Panorama

Workflow	Files	Folder
Qwen-Image-Edit 2511 Image-to-Panorama (image→pano)	`qwen_image_edit_2511_bf16.safetensors` (or fp8mixed), `Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors`, `qwen_2.5_vl_7b_fp8_scaled.safetensors`, `qwen_image_vae.safetensors`	`diffusion_models/`, `loras/`, `text_encoders/`, `vae/`
Qwen-Image 2512 + 360 LoRA (text→pano)	`qwen_image_2512_fp8_e4m3fn.safetensors`, `Qwen-Image-2512-Lightning-4steps-V1.0-fp32.safetensors`, `qwen-360-diffusion-2512-int8-bf16-v2.safetensors`, `qwen_2.5_vl_7b_fp8_scaled.safetensors`, `qwen_image_vae.safetensors`	`diffusion_models/`, `loras/`, `loras/`, `text_encoders/`, `vae/`

360 LoRA download: https://huggingface.co/ProGamerGov/qwen-360-diffusion

Third-party custom node dependencies

A few workflows rely on a third-party ComfyUI custom node:

LaMa Erase → Acly/comfyui-inpaint-nodes
Fooocus SDXL Inpaint / Outpaint → uses the same comfyui-inpaint-nodes plugin for its INPAINT_* nodes.
(Planned, not shipped) Demucs Vocals / Background → lum3on/ComfyUI_AudioTools for the AudioStemSeparate node.

All other workflows use only ComfyUI core nodes.

Not using the shipped workflows?

You don’t need to download any of the above. ComfyTV runs any ComfyUI GUI workflow JSON you drop into workflows/<kind>/, together with a _preset.json that maps stage inputs to its nodes. Build it in regular ComfyUI, save it, drop it in, restart ComfyUI, and your custom option appears in the matching stage’s workflow dropdown. See custom-workflows.md for the _preset.json format and examples.

​Generate · Text

​Generate · Image

​Generate · Video (LTX 2.3)

​Generate · Audio

​Image edits

​Image variations (single-workflow N-output)

​Relight

​Panorama

​Third-party custom node dependencies

​Not using the shipped workflows?

Generate · Text

Generate · Image

Generate · Video (LTX 2.3)

Generate · Audio

Image edits

Image variations (single-workflow N-output)

Relight

Panorama

Third-party custom node dependencies

Not using the shipped workflows?