Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > ComfyUI-GGUF-Loader > Checkpoint Loader (GGUF) ⚡

ComfyUI Node: Checkpoint Loader (GGUF) ⚡

Class Name

CheckpointLoaderGGUF

Category
🤖 CCTech/GGUF
Author
chriscoletech (Account age: 0days)
Extension
ComfyUI-GGUF-Loader
Latest Updated
2026-11-15
Github Stars
0.03K

How to Install ComfyUI-GGUF-Loader

Install this extension via the ComfyUI Manager by searching for ComfyUI-GGUF-Loader
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter ComfyUI-GGUF-Loader in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

Checkpoint Loader (GGUF) ⚡ Description

Loads safetensors or GGUF checkpoints, providing model, CLIP, and VAE when available.

Checkpoint Loader (GGUF) ⚡:

CheckpointLoaderGGUF loads an image-generation checkpoint into the components a ComfyUI workflow commonly needs. It accepts both standard safetensors checkpoints and GGUF files, so you can use one loader for either format. For GGUF files, it uses GGUF-aware operations and keeps quantized model weights in their GGUF representation during loading. It then asks ComfyUI to identify the model architecture and load the components available in the file. This is especially useful for compatible all-in-one checkpoints, where the text-processing and image-decoding components can come from the same file rather than separate loaders. What the node can provide still depends on the checkpoint’s contents and whether ComfyUI recognizes its architecture.

Checkpoint Loader (GGUF) ⚡ Input Parameters:

ckpt_name

Selects the checkpoint file to load. This required dropdown is populated from registered checkpoint folders and GGUF diffusion-model folders; its available choices depend on the files in your installation, so there is no fixed options list. The node description specifies safetensors and GGUF checkpoints under models/checkpoints and GGUF files under models/diffusion_models. Your selection determines which model architecture and components ComfyUI attempts to load, so choose the file intended for your workflow. Minimum: not applicable. Maximum: not applicable. Default: none specified in the source.

Checkpoint Loader (GGUF) ⚡ Output Parameters:

model

A MODEL connection containing the loaded image-generation model. Connect it to a compatible sampling or model-processing node. For a GGUF selection, the node wraps the detected model in a GGUF-aware model patcher so its quantized weights can be handled during loading and patching. This output does not, by itself, establish that the checkpoint also contains usable text or image-decoding components.

clip

A CLIP connection for the checkpoint’s loaded text-processing component, when one is available and recognized. In a compatible workflow, connect it to the text-encoding node used to turn your prompt into conditioning for the model. Some supported all-in-one checkpoints supply this component from the same file. A model-only GGUF should not be assumed to provide it; if this output is absent or unusable, you may need a separate compatible text-encoder loader.

vae

A VAE connection for the checkpoint’s loaded image-decoding component, when one is available and recognized. A compatible downstream node uses it to turn generated image data into an image you can view or save. The node requests a VAE during loading, but its presence and behavior depend on the selected checkpoint. The source notes that SenseNova U1.5 uses a pixel-space dummy VAE from its all-in-one checkpoint; do not assume every GGUF has the same arrangement.

Checkpoint Loader (GGUF) ⚡ Usage Tips:

  • Place the file in a folder registered for the relevant model category, then refresh ComfyUI’s model list if it does not appear in ckpt_name. The dropdown is built from discovered filenames, not from a path you type into the node.
  • For a compatible all-in-one checkpoint such as the SenseNova U1.5 example described in the source, try the node’s model, clip, and vae connections together before adding separate component loaders.
  • For a GGUF containing only diffusion-model weights, check whether the clip and vae connections are usable in your workflow. Use separate compatible component loaders when the file does not provide those components.
  • If you have both safetensors and GGUF versions of a compatible checkpoint, select the version that fits your workflow. The GGUF branch uses GGUF-aware operations and retains quantized weights; the safetensors branch uses ComfyUI’s checkpoint-loading path. This node exposes no separate quantization, device, or performance setting to tune.
  • If a GGUF model needs a supplied transformer-configuration metadata sidecar, keep it beside the model with the expected checkpoint-specific name. The source also permits a folder-level sidecar when it unambiguously identifies that file. Avoid placing unrelated metadata beside a checkpoint and assuming it will apply.

Checkpoint Loader (GGUF) ⚡ Common Errors and Solutions:

ERROR: Could not detect model type of: <path>

  • Explanation: This is an explicit error from the GGUF branch. ComfyUI could not identify a usable model from the loaded file.
  • Solution: Confirm that you selected a supported image-generation GGUF rather than a different kind of GGUF file. Check that the download is complete and that any required, model-specific metadata sidecar is present. If the architecture requires newer loader support, check compatibility between your ComfyUI installation and this custom node.

FileNotFoundError or a checkpoint path that cannot be resolved

  • Explanation: The selected filename no longer resolves to a file in the registered model folders, for example after a move, rename, or removal. The exact message may vary with the underlying file lookup.
  • Solution: Check that the file still exists in its registered folder, refresh the model list, and select it again.

JSONDecodeError while reading a metadata sidecar

  • Explanation: A selected GGUF metadata sidecar contains invalid JSON, or its embedded config string is not valid JSON. The node reads and parses that content when applicable.
  • Solution: Obtain the correct sidecar for the checkpoint or correct its JSON syntax. Keep checkpoint-specific sidecars paired with the file they describe; do not substitute configuration from another model.

Found <number> metadata sidecars claiming <filename>; none named after it, skipping them all.

  • Explanation: This warning means several folder-level metadata files claim the same GGUF, so the node declines to choose among them. Loading may then fail if the model needs that metadata.
  • Solution: Remove or correct the conflicting claims, or provide the intended sidecar using a checkpoint-specific filename beside the GGUF.

size mismatch or Error(s) in loading state_dict

  • Explanation: Model weights do not fit the architecture ComfyUI constructed. One possible cause is missing or incorrect configuration metadata, but the message alone does not prove that is the cause.
  • Solution: Verify that the GGUF and any sidecar belong to the same model and version. Replace a mismatched sidecar with the correct one, confirm the file is complete, and check whether the architecture is supported by your installed loader.

A downstream node reports a missing or incompatible CLIP or VAE

  • Explanation: The loader requests both components, but the selected checkpoint may not contain a usable version of one or both. A model-only GGUF is not necessarily an all-in-one checkpoint.
  • Solution: Use separate text-encoder or VAE loaders appropriate for that model, or select a compatible all-in-one checkpoint. Check the model’s required workflow before connecting components from different files.

Out of memory or a device-memory allocation error

  • Explanation: Loading or using the selected model may exceed available memory. The supplied code also notes that moving weights to release a mapped file can potentially run out of memory; it does not promise a particular memory saving for every GGUF.
  • Solution: Close other memory-intensive tasks, try a smaller compatible checkpoint, or use memory-management settings available elsewhere in your ComfyUI setup. This node has no memory-control input of its own.

Checkpoint Loader (GGUF) ⚡ Related Nodes

Go back to the extension to check out more related nodes.
ComfyUI-GGUF-Loader
RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

Checkpoint Loader (GGUF) ⚡