Model support

Additional models will be added as they become available and there is public interest in them
See models overview for details on each model, including their architecture, complexity and other info

• RunwayML Stable Diffusion 1.x and 2.x (all variants)
• StabilityAI Stable Diffusion XL, StabilityAI Stable Diffusion 3.0 Medium, StabilityAI Stable Diffusion 3.5 Medium, Large, Large Turbo
• StabilityAI Stable Video Diffusion Base, XT 1.0, XT 1.1
• StabilityAI Stable Cascade Full and Lite
• Black Forest Labs FLUX.1 Dev and Schnell
• Black Forest Labs FLUX.1 Fill and Depth
• Black Forest Labs FLUX.1 Kontext-Dev
• Black Forest Labs FLUX.1 Krea-Dev
• lodestones Chroma Standard, Detail Calibrated and Flash
• FreePik F-Lite Standard, Texture and 7B
• NVLabs Sana 1.0 and 1.5
• nVidia Cosmos-Predict2 T2I 2B and 14B
• AuraFlow 0.3 and 0.2
• AlphaVLLM Lumina-Next-SFT
• AlphaVLLM Lumina 2
• Playground AI v1, v2 256, v2 512, v2 1024 and latest v2.5
• Tencent HunyuanDiT 1.2 and 1.1
• OmniGen
• Meissonic
• Kwai Kolors
• Ostris Flex.1-Alpha
• CogView 3+
• LCM: Latent Consistency Models
• aMUSEd 256 and 512
• Segmind Vega, Segmind SSD-1B, Segmind SegMoE SD and SD-XL, Segmind SD Distilled (all variants)
• Kandinsky 2.1, 2.2, 3.0
• PixArt-α XL 2 Medium and Large
• PixArt-Σ
• Warp Wuerstchen
• Tsinghua UniDiffusion
• DeepFloyd IF Medium and Large
• ModelScope T2V
• BLIP-Diffusion
• KOALA 700M
• VGen
• SDXS
• Hyper-SD
• HiDream-I1 Full, Dev, Fast
• Wan 2.1 1.3B and 14B T2V models for T2I workflows
• Wan 2.2 5B and A14B T2V models for T2I workflows
• Bria 3.2
• Qwen-Image
• Qwen-Image-Edit

Support Notes

• HunyuanDit models requires explicit float16 precision
  set in settings -> compute settings -> model precision -> fp16
• aMUSEd models require explicit noise generator device to be unset
  set in settings -> compute settings -> generator device -> unset