Published on August 25, 2026, ComfyUI_VoxCPM_SM offers VoxCPM2 inference and LoRA fine-tuning with custom voice data, with stated support for low VRAM and quantized models.
A publication dated August 25, 2026 presents ComfyUI_VoxCPM_SM, a ComfyUI plugin for VoxCPM2 inference and LoRA fine-tuning using custom voice data. The announcement says it supports low-VRAM use cases and quantized models, including INT4, INT8, and GGUF.
Engineers can use a node-based workflow to run and customize voice generation locally. To confirm compatibility, requirements, and results, consult the original publication and plugin materials; check versions, settings, and behavior in your environment before adopting the workflow. If you use AI to study or apply the material, avoid submitting recordings or identifiable data without authorization and follow your organization’s data policy.