Instant Load Quadruple CLIP

This is the most niche node in the pack, and the pack knows it. Instant Load Quadruple CLIP takes four text encoder dropdowns and fuses them into a single CLIP output. Stock ComfyUI tops out at dual encoders; triple and quadruple loaders only exist in the custom-node world, for experimental multi-encoder recipes that want more text representations than mainstream pipelines use.

If you're here, you almost certainly arrived from a specific workflow that has this node in it. That's fine - that's the use case. But let's be clear about scope: the CLIPType is hardcoded to the stable-diffusion family, there's no recipe dropdown to guide you, and the four files you pick have to be the exact set that workflow expects, in the order it expects. It is not a "throw four encoders at it and see" node.

Inputs and output

• clip_name1, clip_name2, clip_name3, clip_name4 - four dropdowns over ComfyUI/models/text_encoders.
• CLIP - the fused encoder, wired to CLIPTextEncode.

Mechanically it's the same streaming InstantTensor path as the rest of the CLIP family: for .safetensors files the headers are read first, the encoder models are constructed, then tensors stream via copy_() straight into the parameters. With four encoders loading at once that peak-memory saving is the whole reason the node is worth having - a quadruple load is the worst-case spike in this family.

Common issues

All four files need to be .safetensors for the streaming path to engage; anything else falls through to ComfyUI's normal loader. CUDA-only, instanttensor must be installed or you'll get "The instanttensor package is required for CUDA instant loading," and remember the pack is a brand-new single-commit release (May 2026) - there is no community troubleshooting lore to lean on, so verify the four-encoder combo against the source workflow before changing anything.

How to install

ComfyUI Manager → search ComfyUI-InstantTensorLoaders → install and accept the dependency install. Or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/redstonewhite/ComfyUI-InstantTensorLoaders

Restart ComfyUI; it appears under advanced/loaders/instanttensor. If the pip dependency didn't land, run pip install instanttensor in the venv that runs ComfyUI and restart.