koboldcpp

All-in-one local AI server (LLM, image, speech, TTS) on ggml/llama.cpp

KoboldCpp is a single-binary local AI server built on ggml/llama.cpp that bundles LLM text generation (KoboldAI Lite web UI + an OpenAI-compatible API), Stable Diffusion image generation, Whisper speech-to-text and TTS. This ebuild builds the ggml backends from source and installs the pure-Python launcher (koboldcpp.py) with the embedded web UI and resources; it does NOT ship the upstream PyInstaller binary or any model weights. AMD/NVIDIA GPU acceleration is via Vulkan (the officially supported path); ROCm/hipBLAS is best-effort.

homepage ↗ github: LostRuins/koboldcpp