How popular is stable-diffusion.cpp on GitHub?

leejet/stable-diffusion.cpp has 7,498 stars and 846 forks on GitHub, and gained 203 stars in the last 7 days.

What license does stable-diffusion.cpp use?

leejet/stable-diffusion.cpp is released under the MIT license.

leejet/stable-diffusion.cppPublic

Diffusion model(SD,Flux,Wan,Qwen Image,Z-Image,...) inference in pure C/C++

AI summary: A pure C/C++ implementation for running Stable Diffusion and other diffusion models locally without Python dependencies.

Star on GitHub

Stars: 7.5K; +30 today
Forks: 846
Watchers: 78
Open issues: 228
Open PRs: 39
Contributors: ~113
Commits: 929
Branches: 4

C++MITCreated Aug 13, 2023Last push 7d agoLatest release master-929-3f8527a+203 stars this week+267 this month

Quick answers

What is stable-diffusion.cpp?: A pure C/C++ implementation for running Stable Diffusion and other diffusion models locally without Python dependencies.
What does stable-diffusion.cpp do?: Stable-diffusion.cpp provides a lightweight, highly optimized inference engine for generating images from text prompts using diffusion models. By porting the complex PyTorch architectures into raw C/C++, it eliminates massive dependencies and enables execution across diverse hardware environments. It supports a wide array of state-of-the-art models including Flux, Qwen-Image, and various SDXL variants. The project focuses on maximizing performance while minimizing memory footprint, making local AI generation accessible on consumer hardware.
Who is stable-diffusion.cpp for?: C++ developers, AI engineers, and enthusiasts wanting fast, local image generation. Requires basic command-line familiarity to run.
How do I get started with stable-diffusion.cpp?: ./bin/sd -m models/v1-5-pruned-emaonly.safetensors -p "a lovely cat"
How popular is stable-diffusion.cpp on GitHub?: leejet/stable-diffusion.cpp has 7,498 stars and 846 forks on GitHub, and gained 203 stars in the last 7 days.
What license does stable-diffusion.cpp use?: leejet/stable-diffusion.cpp is released under the MIT license.

Star history

since Aug 13, 2023

7.5K stars as of Oct 3, 2026. Before Sep 24, 2026, reconstructed from public GitHub event archives (checked against the repository's real star total); since then measured daily.

Contribution activity

commits per day, last 52 weeks

619 commits in the last yearLessMore

Signals and awards

derived from tracked data

Community-driven
~113 contributors
Permissive license
MIT
Continuous integration
Automated checks passing

What stable-diffusion.cpp does

Stable-diffusion.cpp provides a lightweight, highly optimized inference engine for generating images from text prompts using diffusion models. By porting the complex PyTorch architectures into raw C/C++, it eliminates massive dependencies and enables execution across diverse hardware environments. It supports a wide array of state-of-the-art models including Flux, Qwen-Image, and various SDXL variants. The project focuses on maximizing performance while minimizing memory footprint, making local AI generation accessible on consumer hardware.

C++ developers, AI engineers, and enthusiasts wanting fast, local image generation. Requires basic command-line familiarity to run.

Zero Python dependency: Runs inference completely standalone in C/C++, simplifying deployment and avoiding environment hell.
Broad model support: Continually updated to support the latest diffusion architectures like Flux, SD3, and Wan.
Hardware acceleration: Leverages CPU vector instructions and GPU backends like CUDA, Metal, and Vulkan for maximum throughput.
Low memory overhead: Implements aggressive memory management techniques to fit large models onto consumer GPUs.
Cross-platform compatibility: Compiles and runs natively on Windows, macOS, and Linux without complex configuration.

Where teams use it

Local Image Generation

Run high-quality image synthesis on personal laptops or edge devices without needing a cloud API or heavy PyTorch install.

Application Integration

Embed state-of-the-art diffusion models directly into C++ applications, games, or native software products.

Resource Constrained Inference

Deploy generative AI on older or memory-constrained hardware where standard Python frameworks would fail.

Standalone Binaries

Distribute AI generation capabilities as simple, self-contained executable files.

Getting started: ./bin/sd -m models/v1-5-pruned-emaonly.safetensors -p "a lovely cat"

README

master branch

stable-diffusion.cpp

Diffusion model(SD,Flux,Wan,...) inference in pure C/C++

Note that this project is under active development.
API and command-line option may change frequently.

🔥Important News

2026/09/20 🚀 stable-diffusion.cpp adds Day-0 support for Qwen-Image-2.1
2026/08/20 🚀 stable-diffusion.cpp now supports LTX-2.5
2026/08/04 🚀 stable-diffusion.cpp adds Day-1 support for MiniMax-H3
2026/06/25 🚀 stable-diffusion.cpp now supports Krea2
2026/06/04 🚀 stable-diffusion.cpp now supports Ideogram4
2026/05/31 🚀 stable-diffusion.cpp now supports PiD
2026/05/27 🚀 stable-diffusion.cpp now supports Lens
2026/05/17 🚀 stable-diffusion.cpp now supports LTX-2.3
2026/04/11 🚀 stable-diffusion.cpp now uses a brand-new embedded web UI.
2026/01/18 🚀 stable-diffusion.cpp now supports FLUX.2-klein
2025/12/01 🚀 stable-diffusion.cpp now supports Z-Image
2025/11/30 🚀 stable-diffusion.cpp now supports FLUX.2-dev
2025/10/13 🚀 stable-diffusion.cpp now supports Qwen-Image-Edit / Qwen-Image-Edit 2509
2025/10/12 🚀 stable-diffusion.cpp now supports Qwen-Image
2025/09/14 🚀 stable-diffusion.cpp now supports Wan2.1 Vace
2025/09/06 🚀 stable-diffusion.cpp now supports Wan2.1 / Wan2.2

Features

Plain C/C++ implementation based on ggml, working in the same way as llama.cpp
Super lightweight and without external dependencies
Supported models
- Image Models
- Image Edit Models
- Video Models
- PhotoMaker support.
- IP-Adapter support (SD 1.5 and SDXL, including Plus)
- Control Net support with SD 1.5
- ADetailer
- LoRA support, same as stable-diffusion-webui
- Latent Consistency Models support (LCM/LCM-LoRA)
- Faster and memory efficient latent decoding with TAESD
- Upscale images generated with ESRGAN
Supported backends
- CPU (AVX, AVX2 and AVX512 support for x86 architectures)
- CUDA
- Vulkan
- Metal
- OpenCL
- SYCL
Supported weight formats
- Pytorch checkpoint (.ckpt or .pth or .pt)
- Safetensors (.safetensors)
- GGUF (.gguf)
Convert mode supports converting model weights to .gguf or .safetensors
Supported platforms
- Linux
- Mac OS
- Windows
- Android (via Termux, Local Diffusion)
Flash Attention for memory usage optimization
Negative prompt
stable-diffusion-webui style tokenizer (not all the features, only token weighting for now)
VAE tiling processing for reduce memory usage
Sampling method
- Euler A
- Euler
- Heun
- DPM2
- DPM++ 2M
- DPM++ 2M v2
- DPM++ 2S a
- ER-SDE
- LCM
Cross-platform reproducibility
- --rng cuda, default, consistent with the stable-diffusion-webui GPU RNG
- --rng cpu, consistent with the comfyui RNG
Embedds generation parameters into png output as webui-compatible text string

Quick Start

Get the sd executable

Download pre-built binaries from the releases page
Or build from source by following the build guide

Download model weights

download weights(.ckpt or .safetensors or .gguf). For example
- Stable Diffusion v1.5 from https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5
```
curl -L -O https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors
```

Generate an image with just one command

./bin/sd-cli -m ../models/v1-5-pruned-emaonly.safetensors -p "a lovely cat"

For detailed command-line arguments, check out cli doc.

Performance

If you want to improve performance or reduce VRAM/RAM usage, please refer to performance guide. For runtime and parameter backend placement, see the backend selection guide.

More Guides

Bindings

These projects wrap stable-diffusion.cpp for easier use in other languages/frameworks.

Golang (non-cgo): seasonjs/stable-diffusion
Golang (cgo): Binozo/GoStableDiffusion
Golang (non-cgo): l8bloom/gosd
C#: DarthAffe/StableDiffusion.NET
Python: william-murray1204/stable-diffusion-cpp-python
Rust: newfla/diffusion-rs
Flutter/Dart: rmatif/Local-Diffusion

UIs

These projects use stable-diffusion.cpp as a backend for their image generation.

Contributors

Thank you to all the people who have already contributed to stable-diffusion.cpp!

View on GitHub

Recent activity

commits and pull requests

ci: bundle HIP runtime libraries with Windows ROCm releaseOpen
#2090 by harkgill-amd · updated 2d ago
Fix ROCm ccache build caching and update to ROCm 10.0.0Open
#1890 by superm1 · updated 2d ago
hexagon: add an experimental Qualcomm FP8 DiT and VAE pathOpen
#1970 by happyyzy · updated 2d ago
fix: correct non-circular tile placement and blendingOpen
#2088 by shawn-mengchen-xu · updated 2d ago
feat: run conditional and unconditional CFG in one batched UNet forwardOpen
#2085 by daniandtheweb · updated 2d ago
fix: avoid scanning cwd during server startupOpen
#2086 by mikemikimike · updated 3d ago
Fix signed shape-product wrap in safetensors reader (DoS via NULL deref)Closed
#1875 by x14ngch3n · updated 3d ago
fix: improve GGUF metadata parsing and reader selectionMerged
#2062 by losewayy · updated 3d ago
feat: enable hip INT8 tensorwise matmul and convrot (#2071)
leejet committed 3f8527a · 7d ago
fix: keep scaled INT8 convrot matmuls on Vulkan (#2070)
leejet committed 42ab1c1 · 7d ago
fix: update ggml to fix INT8 convrot backend fallback (#2069)
leejet committed 27f7c43 · 7d ago
fix: load INT8 convrot LLM embeddings correctly (#2067)
leejet committed ede32a6 · 7d ago

Recent open issues

view all

[Feature] ADetailer Server Support
#2087 · 2d ago · by drone540
enhancement
[Bug] --lora-apply-mode immediately on a quantized Qwen-Image-2.1 GGUF applies 0/N LoRA tensors then aborts (Vulkan)
#2083 · 4d ago · by xiaobaoheng-lang · 2 comments
[Bug] sd-server: request accepted during startup window never responds, process spins at high CPU until killed
#2081 · 4d ago · by bbirchman
bug
[Bug] Pervasive Green Artifacts and Latent Corruption with Qwen-Image-2.1 Distillation Accelerated (DMD2) Models and LoRAs
#2078 · 5d ago · by eleven-monkey · 1 comments
bug
[Feature] Generalize video hires / latent spatial upscale beyond LTX (MiniMax-H3)
#2077 · 6d ago · by farawayso
enhancement

Discussions

all 169

Show and tell: Vison, a Windows desktop app built on stable-diffusion.cpp (Vulkan, FLUX / Z-Image / Wan)
Show and tell · 2d ago
How to use the ROCm build ?
Q&A · 10d ago · 2 comments
Diffusion Desk - Desktop App GUI around stable-diffusion.cpp and llama.cpp
Show and tell · 3mo ago · 1 comments
Which area should be prioritized next in development?
Polls · 2mo ago · 3 comments
Built on CUDA 11.5 and run several NVIDIA architectures on 8.6 (Ampere) GPU (runtime compile of virtuals takes minutes, empty images on virtual Volta and below)
Benchmark · 4d ago

Releases and announcements

708 total

master-929-3f8527amaster-929-3f8527aSep 27, 202613.3K downloads
master-928-42ab1c1master-928-42ab1c1Sep 27, 2026310 downloads
master-927-27f7c43master-927-27f7c43Sep 27, 2026188 downloads
master-926-ede32a6master-926-ede32a6Sep 27, 2026631 downloads
master-925-47e83d7master-925-47e83d7Sep 27, 2026172 downloads

Commits per week

last 52 weeks

619 commits in the last 52 weeks.

When work happens

weekday and hour

Commit volume by weekday and hour (UTC). Larger dots mean more commits.

Who is committing

last 52 weeks

Maintainer commits311 (50%)

Community commits309 (50%)

620 commits in total over the last year.

Date	List	Rank	Stars gained
Sep 25, 2026	daily	#12	+33
Sep 24, 2026	daily	#12	+33

openclaw/openclaw
The AI that really does things. Any OS. Any Platform. The lobster way. 🦞
391.3K stars · TypeScript
obra/superpowers
An agentic skills framework & software development methodology that works.
295.2K stars · Shell
NousResearch/hermes-agent
The agent that grows with you
251.2K stars · Python
n8n-io/n8n
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
206.7K stars · TypeScript
tensorflow/tensorflow
An Open Source Machine Learning Framework for Everyone
200.7K stars · C++
firecrawl/firecrawl
Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥
188.6K stars · TypeScript