Saltar al contenido
Node packVerificadoExterno
Abrir fuente

comfyui-ea-nodes

Custom ComfyUI nodes for advanced video & image workflows — integrating seamlessly with WAN, Flux, and more.

Responsable@ExoticArtsUnratedcalidad

Perfil del paquete

Estado
Verificado
Publicador
@ExoticArts
Fuente
GitHub

Léeme

CI

ComfyUI EA Nodes

Compact, user-friendly LoRA stacking nodes for ComfyUI. Designed to work cleanly with WAN 2.2 video workflows and with SD/SDXL/Flux text-encoder pipelines.


What’s inside

  • EA Power LoRA — stack multiple LoRAs onto the MODEL only (no CLIP).
  • EA Power LoRA +CLIP — stack LoRAs onto MODEL and CLIP.
  • EA Power LoRA WanVideo — stack LoRAs using WanVideoWrapper style sockets (prev_lora / blocks → lora) so WAN 2.2 graphs stay compact.
  • EA Trim Frames — trim frames from the start/end of an image sequence and return first/last previews + frame count.
  • EA Filename → Combine — build a friendly filename prefix and full-path stub (with an optional “trigger” string to avoid cache collisions).

All LoRA pickers list files from ComfyUI/models/loras (falls back to a text box if the list can’t be fetched). Missing LoRAs are safely ignored at runtime.


Nodes

EA Power LoRA

Apply one or more LoRAs to the MODEL only (no CLIP path).

Use this when: your pipeline has no text-encoder/CLIP path (e.g., WAN 2.2 UNet-only branches).

Wiring


\[Model Loader] → (model) → \[EA Power LoRA] → (model) → \[…]


EA Power LoRA +CLIP

Apply one or more LoRAs to MODEL and CLIP.

Use this when: your pipeline includes a text encoder (e.g., SD/SDXL/Flux).
To make a LoRA model-only, set its CLIP weight to 0.00.

Wiring


\[Checkpoint Loader] → (model, clip) → \[EA Power LoRA +CLIP] → (model, clip) → \[…]


EA Power LoRA WanVideo

A WanVideo-compatible variant that takes prev_lora and blocks, and outputs a stacked lora. Combine many LoRAs in a single node to keep WAN 2.2 graphs tidy.

Use this when: you’re wiring WanVideoWrapper nodes and want stackable LoRAs without a long chain of single-LoRA nodes.

Wiring (typical WAN 2.2)


\[WanVideo Lora Select] → (prev\_lora, blocks) → \[EA Power LoRA WanVideo] → (lora) → \[WanVideo Set LoRAs]

You can place separate WanVideo stacks for “high” and “low” groups just like you would with the native nodes.


UI at a glance

  • + Add LoRA — add a row.
  • ✓ — enable/disable a row.
  • LoRA — file from /models/loras (or type a name).
  • Weight — UNet/model weight (left-aligned, two decimals).
  • CLIP — text-encoder weight (shown only on +CLIP).
  • × — remove the row.

The header row appears only when you have at least one LoRA.


Notes & tips

  • Which node when
    • WAN 2.2: use EA Power LoRA (or EA Power LoRA WanVideo if you’re using WanVideoWrapper blocks).
    • SD/SDXL/Flux: use EA Power LoRA +CLIP.
  • Order of LoRAs
    LoRA application is additive; order does not change the result.
  • Trigger words
    Many style LoRAs work fine with CLIP low/off; character/concept LoRAs often benefit from some CLIP weight.
  • Performance
    Each LoRA adds memory/compute overhead. If you stack many, consider reducing precision or VRAM usage elsewhere.

Advanced: JSON payload (saved in workflows)

Each node serializes its rows to a hidden loras_json field. You normally don’t need this, but it’s useful for debugging or scripting.

MODEL-only / WanVideo

{
  "rows": [
    { "enabled": true, "name": "foo.safetensors", "strength_model": 1.00 }
  ]
}

MODEL + CLIP

{
  "rows": [
    { "enabled": true, "name": "bar.safetensors", "strength_model": 1.00, "strength_clip": 1.00 }
  ]
}

Unknown names are ignored at execution time.


Utility nodes

EA Trim Frames

Remove frames from the start/end of an image sequence.

Inputs

  • images (IMAGE)
  • skip_first (INT)
  • skip_last (INT)

Outputs

  • images (trimmed sequence)
  • `first_f

Nodos de este paquete

16 nodos
  • EA_AutoTrimPingPong
  • EA_FilenameCombine
  • EA_LightningMotionBias
  • EA_ListVideos
  • EA_ManifestIndex
  • EA_PingPong
  • EA_PowerLora
  • EA_PowerLora_CLIP
  • EA_PowerLora_WanVideo
  • EA_TrimFrames
  • EA_TrimWindow
  • EA_VideoLoad
  • EA_VideoSaveIdempotent
  • EAImageCompare
  • EAImageCompare3Way
  • EAImageCompare4Way

Fuentes

3 fuentes
GitHubFuenteVersión f9ac3ce1a19eae26854372d674124cbba6d5fd18
github.com/ExoticArts/comfyui-ea-nodes

Verificado 5 oct

GitHubFuente
github.com/ExoticArts/comfyui-ea-nodes

Verificado 2 nov

Extractos de fuentes

1 extracto

Source context: Repo ExoticArts/comfyui-ea-nodes