KSampler Matrix Lab compares multiple samplers and schedulers in one grid

Btitkin’s new ComfyUI custom node, KSampler Matrix Lab, solves a common pain point for generative AI artists and developers: efficiently comparing different sampler and scheduler combinations. Instead of manually duplicating KSampler nodes and changing settings one by one, this node runs all selected samplers against all selected schedulers in a single pass. The output is a tidy grid where rows represent samplers and columns represent schedulers, each cell showing the generated result for that specific pairing. The node supports dropdown selections for samplers and schedulers, a common seed for reproducibility or an incrementing seed per cell to explore variation, and per-cell labels for quick identification. If a combination fails (e.g., incompatible parameters), the node displays an error cell rather than crashing the workflow.

The real value lies in speed and clarity. Users can define model, VAE, CLIP, steps, CFG scale, and denoise strength in a unified header, then instantly see how changing the sampler or scheduler affects output. This is especially useful for fine‑tuning diffusion settings, testing new sampler presets, or documenting pipeline behavior. The node is open‑source on GitHub under an MIT license, and btitkin welcomes feedback and feature suggestions. For ComfyUI power users, KSampler Matrix Lab reduces experimentation time from minutes to seconds while maintaining full control over seeding and output formatting.