basicIn_clip

Stock ComfyUI makes you drop two separate "CLIP Text Encode" nodes for a single generation - one for your positive prompt, one for negative - each with its own text box, each needing the same CLIP wired in twice. basicIn_clip folds that into one node: one CLIP input, two text boxes, two CONDITIONING outputs. It's a small thing, but it's exactly the kind of clutter-reduction this whole pack is built around - the README opens by explicitly naming dense, hard-to-read workflow graphs as the problem it exists to solve.

What it is and why you'd reach for it

Every txt2img or img2img generation needs both a positive and a negative conditioning, and in a stock ComfyUI graph that's two nodes, two wires from your CLIP loader, and two more wires headed into your sampler. basicIn_clip collapses all of that into a single box, which matters more than it sounds like once you're building or reading a graph with several sampling stages - fewer nodes to scan for, fewer wires crossing the canvas, one place to check when a prompt isn't behaving the way you expect.

The inputs and outputs that matter

Three inputs: clip (wire in your loaded CLIP model), positive (a multiline text box - your actual prompt), and negative (single-line). Two outputs, positive and negative, both CONDITIONING - wire them straight into a sampler's positive and negative slots exactly as you would from two separate encode nodes.

One thing worth knowing before you fill in negative: it's genuinely a no-op on a large and growing share of current models. Guidance-distilled models - Z-Image Turbo, Flux 2 Klein's distilled variants, and most of what's mainstream in 2026 - run at CFG 1 by default, and at CFG 1 there's no unconditional pass for a negative prompt to steer. The field still exists here and still takes text; it just does nothing until you're on a model (or CFG setting) where negative conditioning actually has something to attach to.

Installing it

Search "ComfyUI-Apt_Preset" in ComfyUI Manager, or:

cd ComfyUI/custom_nodes
git clone https://github.com/cardenluo/ComfyUI-Apt_Preset.git

Run install.bat on Windows (pip install -r requirements.txt on Linux/Mac), restart. No model downloads of its own - it's a thin wrapper around ComfyUI's own text-encoding call.

Common issues

If your negative prompt seems to be doing absolutely nothing, don't assume the node is broken - check what model and CFG you're actually running first. On a guidance-distilled model at its default CFG, negative prompts are structurally inert everywhere in ComfyUI, not just here; restating constraints positively in the positive prompt, or adding a dedicated attention-guidance node, is the real fix, not swapping encode nodes. Beyond that, the usual CLIP gotcha still applies: this node doesn't load anything itself, so if clip comes up empty, the problem is upstream in whatever loader you're feeding it from, not in this node.