Sampler Combo

Sampler Combo is the "collect all sampling settings into one node" move, in its full form: sampler, scheduler, steps, and CFG in a single node that hands all four to your KSampler. It's the sibling of the pack's compact version - same idea, but with wider parameter ranges, full-length parameter names, and a bit more headroom for the experimental end of things.

How it works

The dropdowns are generated from comfy.samplers.KSampler at load time, so they always reflect exactly what your ComfyUI build supports - 44 samplers and 9 schedulers, including the newer _cfg_pp sampler variants. Behind the scenes the node validates your selection (sampler exists, scheduler exists, steps and CFG in range) and falls back to safe defaults (euler / normal / 20 / 7.0) rather than crashing on a bad config. That's the whole mechanism: it's a pass-through with a seatbelt, not a smarter sampler.

The practical benefit is workflow tidiness. Instead of four separate widgets you have to hunt for every time you want to change the denoise recipe, everything's in one panel. And because the outputs are plain values (sampler_name, scheduler, steps, cfg), you can wire them anywhere - into KSampler, into a model-sampling node, into a comparison workflow where you're sweeping samplers.

Inputs that matter

• sampler_name - the algorithm. Default euler. For SD 1.5/SDXL the safe all-rounder is DPM++ 2M Karras; for flow-matching models (Flux and the 2026 guidance-distilled line) stay with euler and skip Karras.
• scheduler - the noise schedule, default normal. This is the field people forget they changed when an image suddenly degrades.
• steps - 1 to 100 here (the compact node caps at 50), default 20.
• cfg - 0 to 20, default 7.0. Remember CFG is architecture knowledge: SDXL wants 4–9, distilled models want ~1.

Outputs: sampler_name, scheduler, steps, cfg.

Compact vs. full

If you've seen Sampler Combo (Compact) in the same pack: this is the roomier one. The compact node has tighter ranges (steps ≤ 50, CFG ≤ 15) and shorter param names; this one allows up to 100 steps and CFG 20, which matters if you're doing high-step quality passes or deliberately pushing CFG high on SDXL-lineage models. If you never exceed 50 steps and CFG 15, save the screen space and grab the compact. Otherwise, this is the one.

Installing it

ComfyUI-KikoTools pack: ComfyUI Manager → "ComfyUI-KikoTools" → Install → restart, under ComfyAssets → Samplers. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/ComfyAssets/ComfyUI-KikoTools.git
cd ComfyUI-KikoTools
pip install -r requirements.txt

Restart ComfyUI.

Gotchas

The node won't stop you from picking a schedule that fights your model - the seatbelt is for invalid configs, not bad taste. The classic is grabbing a workflow with karras set and running it on a flow-matching model, where Karras distorts the trajectory; if output quality collapses after "just tweaking the sampler," check the scheduler first. And because the dropdowns come from your running ComfyUI, a ComfyUI update can rename or remove a sampler - old workflows occasionally need their sampler re-picked from the new list.