This is<a target="_blank" rel="ugc" href="https://huggingface.co/city96/Qwen-Image-gguf/tree/main"> city96</a>'s quantized versions of of Qwen-Image checkpoints

Qwen-Image GGUF's from City96

This is<a target="_blank" rel="ugc" href="https://huggingface.co/city96/Qwen-Image-gguf/tree/main"> city96</a>'s Q3_K_M.gguf of Qwen-Image

qwenImageGgufsFrom_q3KMGguf.gguf

Q3_K_M.gguf

Q5_K_S gguf

Q4 GGUF

Q6_K.gguf

Q3_K_S.gguf

Q8 GGUF

Q2_K.gguf

This is a 3-bit quantized GGUF conversion of the <a target="_blank" rel="ugc" href="https://huggingface.co/Qwen/Qwen-Image">Qwen/Qwen-Image</a> model, released by <a target="_blank" rel="ugc" href="https://huggingface.co/city96/Qwen-Image-gguf">City96</a> and mirrored here for convenience. The Q3_K_M variant is optimized for GPUs with at least 10GB VRAM,<ul><li>Tested VRAM Usage:<ul><li>Ubuntu, Firefox (8 tabs): ~8.8 - 9.2 GB VRAM (CFG 3, 20 Steps, uni_pc, normal, 5.4s/it)</li><li>Windows 11, Brave (1 tab), MiniConda, GGUF excluded from Windows Defender: ~9.6 GB VRAM</li></ul>Tips: Offload monitor tasks to an integrated GPU to free up VRAM. Runs smoothly on Linux or Windows</li></ul>

Qwen-Image GGUF  3Q_K_M - 10G VRAM

qwenImageGGUF3QKM10G_q3KMGguf.gguf

Text Encoder

Files with SHA256: ff96f80b90f8234e498c803965857d7850f89011dde805f83f1e80aa741bcdfb