Model
Trends
.ai
Model
Trends
.ai
the best open source ai models

Efficient Loader and KSampler (Efficient) node in ComfyUI

Efficient Loader combines checkpoint, VAE, LoRA stack, both prompts and the empty latent into one node, and KSampler (Efficient) adds live previews, a built-in VAE decode and XY Plot support for comparing settings.
Node pack: efficiency-nodes-comfyui
Category: Popular custom node packs

What Efficient Loader and KSampler (Efficient) does

The Efficiency Nodes pack (by jags111, continuing LucianoCirino's work) compresses the standard six-node text-to-image head into one Efficient Loader: ckpt_name, vae_name, clip_skip, lora_name with strengths (or a LoRA Stack input), positive and negative text, the latent size and batch. Its DEPENDENCIES output bundles everything for the KSampler (Efficient), which can decode and preview inline and saves a step of wiring.
The XY Plot node is the pack's signature feature: feed it to KSampler (Efficient) to render a grid across seeds, steps, cfg, samplers, schedulers, checkpoints or LoRA strengths, the ComfyUI version of the A1111 X/Y/Z plot. HighRes-Fix Script adds a hires pass to the sampler, and the Noise Control Script tweaks noise for variations.

Inputs

Name
Type
What it is
ckpt_name / vae_name / clip_skip
COMBO / INT
Model, VAE and clip skip (-1 default, -2 for anime models).
lora_name / lora_model_strength / lora_clip_strength
COMBO / FLOAT
One LoRA, or connect a LoRA Stack.
positive / negative
STRING
Prompts typed directly into the node.
empty_latent_width / height / batch_size
INT
The starting latent.
lora_stack / cnet_stack
LORA_STACK / CONTROL_NET_STACK
Optional stackers.

Outputs

Name
Type
What it is
MODEL / CONDITIONING+ / CONDITIONING- / LATENT / VAE / CLIP
various
The usual objects.
DEPENDENCIES
DEPENDENCIES
Bundle for KSampler (Efficient) and XY Plot.
AD
Skip the setup: ComfyUI workflows preinstalled on the cloud
BitVector runs ready-made ComfyUI workflows on its own GPUs. No install, no missing nodes, no red boxes. Open it on your phone or laptop and generate in a minute.

How to use Efficient Loader and KSampler (Efficient)

  1. Install efficiency-nodes-comfyui from Manager.
  2. Add Efficient Loader, pick the checkpoint and type the prompts.
  3. Add KSampler (Efficient); connect model, positive, negative, latent and optional vae; set preview_method and vae_decode true.
  4. Add Save Image to the IMAGE output.
  5. For a grid, add XY Plot with X: steps, Y: cfg and connect it to the sampler script input.

Settings and tips

  • vae_decode "true (tiled)" saves VRAM on large latents.
  • XY Plot with LoRA strengths 0.4/0.6/0.8/1.0 shows you where a LoRA breaks.
  • Set the font size and grid spacing in the plot node when labels overlap.
  • The Efficient Loader prompts are plain text: use the core CLIP Text Encode when you need weights from other nodes.

Troubleshooting Efficient Loader and KSampler (Efficient)

The pack fails to import after a ComfyUI update

Why it happens
Efficiency Nodes has lagged behind core changes several times.

How to fix it
Update the pack through Manager; if it still fails, check the GitHub issues for a patched fork.

XY Plot produces a single image

Why it happens
The plot node is not connected to the script input of KSampler (Efficient), or the X/Y inputs are empty.

How to fix it
Connect XY Plot to the sampler script input and give it XY Input nodes (XY Input: Steps and so on).

Prompt weights like (word:1.3) behave differently

Why it happens
Efficient Loader uses the same encoder but the token_normalization and weight_interpretation options differ from A1111 defaults.

How to fix it
Set weight_interpretation to A1111 for familiar behaviour.

AD
Run this workflow from your phone
Every workflow on this page is preinstalled on BitVector with the models and custom nodes already in place. Pick one, type a prompt, done. Zero setup, nothing to download.

Questions about Efficient Loader and KSampler (Efficient)

Is Efficient Loader faster?

No, it saves wiring, not compute. The speed is identical to the core nodes.

Can I use it with Flux?

Partly; it was designed around SD checkpoints. For Flux use the core loaders or rgthree.

Related nodes

Load Checkpoint
ComfyUI core
Load Checkpoint (CheckpointLoaderSimple) opens a .safetensors or .ckpt model file and hands out the three parts every workflow needs: the diffusion MODEL, the CLIP text encoder and the VAE.
KSampler
ComfyUI core
KSampler runs the denoising loop: it takes the model, prompts and a latent and produces the finished latent image, controlled by seed, steps, cfg, sampler, scheduler and denoise.
Power Lora Loader (rgthree)
rgthree-comfy
Power Lora Loader stacks any number of LoRAs in one node with per-row toggles and strengths, replacing chains of Load LoRA nodes; the rgthree pack also adds Fast Groups Bypasser, Context nodes and the Image Comparer.
Load LoRA
ComfyUI core
Load LoRA (LoraLoader) applies a LoRA file to the MODEL and CLIP with separate strengths, so a style, character or concept can be added to any checkpoint without merging.
ComfyUI Manager
ComfyUI-Manager
ComfyUI Manager is the extension that installs, updates and fixes custom node packs and models from inside the interface, resolves missing nodes in imported workflows and snapshots your setup.
AD
Tired of fixing nodes? Let the cloud do it
BitVector keeps hundreds of ComfyUI workflows installed, updated and tested on fast cloud GPUs. No Python, no CUDA errors, no VRAM limits. Works from any browser.

More ComfyUI nodes

CLIP Text Encode (Prompt)
ComfyUI core
CLIP Text Encode turns a text prompt into CONDITIONING using the model text encoder. One node holds the positive prompt, a second one the negative prompt.
Save Image
ComfyUI core
Save Image writes the IMAGE tensor to ComfyUI/output as a PNG, with the whole workflow embedded in the file metadata so the picture can be dragged back into ComfyUI to restore the graph.
VAE Decode
ComfyUI core
VAE Decode converts the sampled LATENT into a pixel IMAGE with the VAE; VAE Decode (Tiled) does the same in tiles for very large images.
Empty Latent Image
ComfyUI core
Empty Latent Image creates the blank latent canvas (width, height, batch_size) that text-to-image sampling starts from; dimensions must be multiples of 8 and match the model family.
Load Image
ComfyUI core
Load Image reads a picture from the ComfyUI input folder (or an upload) and outputs an IMAGE tensor plus a MASK taken from the alpha channel, the starting point for img2img, inpainting, ControlNet and IP-Adapter graphs.
Load Diffusion Model
ComfyUI core
Load Diffusion Model (UNETLoader) loads a bare denoising network such as Flux, SD3.5, Wan or HunyuanVideo from models/diffusion_models, with a weight_dtype option for fp8 to save VRAM.