Best Open-Source Image Generation Models
From Stable Diffusion 1.5 to Flux to the latest video-capable models — open-source image generation has multiplied. The list below covers what you can actually run locally, with VRAM requirements that vary wildly between models.
- 1
Stable Diffusion XL (CoreML)
3.5B paramsHigher quality image generation. CoreML optimized for iOS. Requires 6GB+ usable memory (iPad/Mac recommended).
Min VRAM: 3.34GBQuant: CoreMLSize: 2.843GBLicense: creativeml-openrail-m - 2
Stable Diffusion 1.5 (CoreML)
0.86B paramsClassic image generation model. Pre-converted to CoreML for iOS/Mac. Downloads as zip, auto-extracts.
Min VRAM: 2.5GBQuant: CoreML-PalettizedSize: 1.46GBLicense: creativeml-openrail-m - 3
SDXL Turbo (GGUF)
3.5B paramsSingle-step SDXL. Near-instant image generation.
Min VRAM: 5GBQuant: Q5_0Size: 3.5GBLicense: stability-community - 4
FLUX.1 Schnell (GGUF)
12B paramsFast 1-4 step generation. State-of-the-art quality. Needs 16GB+ RAM.
Min VRAM: 14GBQuant: Q5_0Size: 12GBLicense: apache-2.0 - 5
FLUX.1 Dev (GGUF)
12B paramsHighest quality FLUX model. 20-50 steps. Mac with 24GB+ RAM.
Min VRAM: 14GBQuant: Q5_0Size: 12GBLicense: flux-1-dev-non-commercial - 6
Stable Diffusion 3 Medium (GGUF)
2.5B paramsSD 3 with MMDiT architecture. Superior text rendering.
Min VRAM: 9.15GBQuant: Q8_0Size: 8.653GBLicense: stability-community - 7
Stable Diffusion 1.5 (GGUF)
0.86B paramsSD 1.5 in single-file GGUF format. Alternative to CoreML. Uses stable-diffusion.cpp with Metal acceleration.
Min VRAM: 2.13GBQuant: Q4_0Size: 1.627GBLicense: creativeml-openrail-m - 8
Stable Diffusion 2.1 Base (CoreML)
0.86B paramsSmallest CoreML image generation model. Palettized for minimal size (1.14GB). Runs on any iPhone with 6GB RAM. Default image generation model.
Min VRAM: 1.56GBQuant: CoreML-PalettizedSize: 1.063GBLicense: creativeml-openrail-m - 9
Stable Diffusion 2.1 (GGUF)
0.86B paramsSD 2.1 in GGUF format. Better quality than 1.5.
Min VRAM: 2.66GBQuant: Q8_0Size: 2.163GBLicense: creativeml-openrail-m