leejet/stable-diffusion.cpp
Source: https://github.com/leejet/stable-diffusion.cpp
Structured fields
{
"archived": false,
"checked_at": "2026-10-17T00:00:00.000Z",
"usefulness": "Dependency-free C/C++ (ggml) inference for diffusion models with GGUF quantization and CPU, CUDA, Vulkan, Metal, OpenCL and SYCL backends; supports SD, SDXL, FLUX, Qwen-Image, Z-Image and Wan/LTX video. A route to run image/video models without PyTorch.",
"source_type": "repository",
"availability": "available",
"next_check_at": "2026-10-18T00:00:00.000Z",
"known_limitations": [
"README says ControlNet is supported for SD 1.5 only."
],
"upstream_updated_at": "2026-10-17T06:14:42.000Z"
}Evidence claims
- [upstream_claim] README: diffusion model inference in pure C/C++ built on ggml, supporting SD1.x/2.x/SDXL/SD3, FLUX.1/FLUX.2, Qwen Image, Z-Image, Wan2.1/2.2 and others; image-edit models; LoRA, IP-Adapter, PhotoMaker, ESRGAN upscaling; ControlNet for SD 1.5 only. source
- [upstream_claim] Backends listed: CPU (AVX/AVX2/AVX512), CUDA, Vulkan, Metal, OpenCL, SYCL; loads .ckpt/.pth/.pt, safetensors and GGUF, and its convert mode writes .gguf or .safetensors. source
- [upstream_claim] Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| code | MIT | permitted | permitted | permitted | unknown | Repository page states MIT. |
| model_weights | unknown | unknown | unknown | unknown | unknown | Loads third-party weights; each carries its own license. |