Local image generation models
Image models you can run on your own GPU or Mac, with memory needs at each precision.
| Model | Made by | Backbone | Native output | Quality tier | Released |
|---|---|---|---|---|---|
| FLUX.2 [dev] | Black Forest Labs | 32.2B | 1024×1024 | 92 | 2025-11 |
| HunyuanImage 3.0 (80B-A13B) | Tencent | 83.0B | 1024×1024 | 90 | 2025-09 |
| Qwen-Image 2512 | Alibaba | 20.4B | 1328×1328 | 90 | 2025-12 |
| Qwen-Image | Alibaba | 20.4B | 1328×1328 | 88 | 2025-08 |
| Qwen-Image-Edit 2511 | Alibaba | 20.4B | 1024×1024 | 86 | 2025-12 |
| FLUX.2 [klein] 9B | Black Forest Labs | 9.1B | 1024×1024 | 84 | 2026-01 |
| HiDream-I1 Full | HiDream.ai | 17.1B | 1024×1024 | 84 | 2025-04 |
| Z-Image Turbo | Alibaba (Tongyi) | 6.2B | 1024×1024 | 84 | 2025-11 |
| FLUX.1 Kontext [dev] | Black Forest Labs | 11.9B | 1024×1024 | 82 | 2025-05 |
| FLUX.1 Krea [dev] | Black Forest Labs | 11.9B | 1024×1024 | 82 | 2025-07 |
| Z-Image | Alibaba (Tongyi) | 6.2B | 1024×1024 | 82 | 2026-01 |
| FLUX.1 [dev] | Black Forest Labs | 11.9B | 1024×1024 | 80 | 2024-08 |
| HiDream-I1 Dev | HiDream.ai | 17.1B | 1024×1024 | 80 | 2025-04 |
| Chroma1-HD | Lodestone Rock | 8.9B | 1024×1024 | 78 | 2025-08 |
| FLUX.2 [klein] 4B | Black Forest Labs | 3.9B | 1024×1024 | 78 | 2026-01 |
| FLUX.1 Fill [dev] | Black Forest Labs | 11.9B | 1024×1024 | 76 | 2024-11 |
| HiDream-I1 Fast | HiDream.ai | 17.1B | 1024×1024 | 76 | 2025-04 |
| CogView4 6B | Zhipu AI | 6.4B | 1024×1024 | 70 | 2025-03 |
| FLUX.1 [schnell] | Black Forest Labs | 11.9B | 1024×1024 | 70 | 2024-08 |
| OmniGen2 | BAAI (VectorSpaceLab) | 4.0B | 1024×1024 | 66 | 2025-06 |
| Stable Diffusion 3.5 Large | Stability AI | 8.1B | 1024×1024 | 66 | 2024-10 |
| Juggernaut XL v9 | RunDiffusion | 2.6B | 1024×1024 | 62 | 2024-02 |
| Lumina-Image 2.0 | Shanghai AI Lab | 2.6B | 1024×1024 | 62 | 2025-01 |
| RealVisXL V5.0 | SG161222 | 2.6B | 1024×1024 | 62 | 2024-08 |
| Stable Diffusion 3.5 Large Turbo | Stability AI | 8.1B | 1024×1024 | 62 | 2024-10 |
| Animagine XL 4.0 | Cagliostro Lab | 2.6B | 1024×1024 | 60 | 2025-01 |
| Illustrious XL v0.1 | OnomaAI | 2.6B | 1024×1024 | 60 | 2024-09 |
| Kolors | Kuaishou | 2.6B | 1024×1024 | 60 | 2024-07 |
| Playground v2.5 | Playground AI | 2.6B | 1024×1024 | 60 | 2024-02 |
| AuraFlow v0.3 | fal | 6.9B | 1024×1024 | 58 | 2024-08 |
| DreamShaper XL v2 Turbo | Lykon | 2.6B | 1024×1024 | 58 | 2024-02 |
| Pony Diffusion V6 XL | PurpleSmartAI | 2.6B | 1024×1024 | 58 | 2024-01 |
| Stable Diffusion 3.5 Medium | Stability AI | 2.5B | 1024×1024 | 58 | 2024-10 |
| Animagine XL 3.1 | Cagliostro Lab | 2.6B | 1024×1024 | 56 | 2024-03 |
| Stable Diffusion XL 1.0 | Stability AI | 2.6B | 1024×1024 | 55 | 2023-07 |
| SDXL-Lightning (4-step) | ByteDance | 2.6B | 1024×1024 | 54 | 2024-02 |
| Hunyuan-DiT v1.2 | Tencent | 1.5B | 1024×1024 | 52 | 2024-07 |
| Stable Cascade | Stability AI | 5.2B | 1024×1024 | 52 | 2024-02 |
| Stable Diffusion 3 Medium | Stability AI | 2.0B | 1024×1024 | 52 | 2024-06 |
| PixArt-Σ XL 1024 | PixArt | 0.6B | 1024×1024 | 50 | 2024-04 |
| Kandinsky 3 | Sber AI | 3.0B | 1024×1024 | 48 | 2023-11 |
| Realistic Vision V5.1 | SG161222 | 0.9B | 512×512 | 42 | 2023-07 |
| DreamShaper 8 | Lykon | 0.9B | 512×512 | 40 | 2023-08 |
| SDXL Turbo | Stability AI | 2.6B | 512×512 | 40 | 2023-11 |
| Stable Diffusion 1.5 | Stability AI / Runway | 0.9B | 512×512 | 35 | 2022-10 |
| LCM DreamShaper v7 | Tsinghua (LCM) | 0.9B | 512×512 | 30 | 2023-10 |
| SD Turbo | Stability AI | 0.9B | 512×512 | 28 | 2023-11 |
Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.