Stable Diffusion v1 5 GGUF Model Card Quantized versions of stable diffusion v1 5 in GGUF format for use with stable diffusion.cpp. At the time of publishing, no ready made GGUF weights for SD1.5 were available for sd.cpp runtime — so here we are. Sample generation: "A lovely cat" · seed 357925 · Q8 0 · 512×512 Available Quantizations File Quantization Description v1 5 pruned bf16.gguf BF16 Near lossless, largest file ⚠️ See note below v1 5 pruned Q8 0.gguf Q8 0 High quality, ~half the size of bf16 v1 5 pruned Q4 K.gguf Q4 K Balanced quality/size v1 5 pruned Q4 0.gguf Q4 0 Smallest, fastest, slight quality loss ⚠️ BF16 known issue: The bf16 variant produces a black image when used with SD 1.5 in stable diffusion.cpp. This appears to be a bug specific to SD 1.5 — other model architectures are not affected. Use Q8 0 if you need maximum quality without quantization loss. Quick Start 1. Download the model 2. Build stable diffusion.cpp Requirements: CUDA capable GPU, CMake ≥ 3.18, CUDA Toolkit This was tested on commit d950627 (version master 520 d950627 ). Check your version with: 3. Start the server The server exposes an OpenAI compatible /v1/images/generations endpoint. 4. Generate a…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy