ComfyUI Node
🚀 Flux KV Cache (GGUF)
Enables KV Cache optimization for reference images on Flux family models loaded via GGUF. The cache persists across model offload/reload cycles, preventing the performance degradation seen with the standard Flux KV Cache node on GGUF models.
🚀 Flux KV Cache (GGUF)
- model
- MODEL
◄cache_on_cputrue►
Categorybootleg
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| cache_on_cpuopt | BOOLEAN | true | Store KV cache on CPU (survives VRAM pressure) or GPU (faster but may be lost during offloading) |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |