Originally Posted: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image_turbo">https://huggingface.co/Comfy-Org/z_image_turbo</a><h3 id="z-image-turbo-nvfp4-the-speed-of-light-meets-high-fidelity-realism">🚀 Z-Image Turbo [NVFP4] – The Speed of Light Meets High-Fidelity Realism</h3>Welcome to the next generation of efficiency. This is a specialized quantization of Z-Image Turbo using the cutting-edge NVFP4 (4-bit Floating Point) format. Optimized specifically for NVIDIA Blackwell (RTX 50-series) architecture, this model delivers unmatched inference speeds without sacrificing the photorealistic soul of the original weights.Why NVFP4?Unlike standard 4-bit quantizations, NVFP4 utilizes a dual-level scaling strategy (FP8 micro-blocks + FP32 tensor scaling). This minimizes quantization errors, preserving the fine textures, skin details, and lighting accuracy that define high-end diffusion models, all while slashing VRAM usage.<h3 id="key-features">✨ Key Features</h3><ul><li>Insane Speed: Generate 1.5MP to 2MP images in seconds.</li><li>VRAM Efficient: Extremely lightweight footprint (approx. 4.2GB), making it accessible for GPUs with lower memory while flying on high-end cards.</li><li>Zero-CFG Distilled: Designed for 8-10 step generations with no CFG overhead.</li><li>Bilingual Mastery: Exceptional at rendering both English and Chinese text within the image, also Spanish</li><li>Raw Realism: Maintains the "organic" look—perfect for those who hate the "AI plastic" aesthetic.</li></ul><h3 id="recommended-settings">🛠 Recommended Settings</h3><ul><li>Steps: 8 to 12 (10 is the sweet spot).</li><li>CFG Scale: 1.0 (The model is distilled; higher values are not needed).</li><li>Sampler: Euler a &amp; bong tangent.</li><li>Resolution: Optimized for 1024x1024 and 1536x1536.</li><li>Architecture: Works best in environments supporting CUDA 13 and Forge UI.</li></ul>Pro Tip: For the best results, use descriptive, natural language prompts. This model excels at "unposed" photography, cinematic lighting, and complex material textures like weathered skin or textile weaves.

z_image_turbo_nvfp4

zImageTurboNvfp4_v10.safetensors

zImageTurboNvfp4_v10_txt.safetensors

v1.0

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:<ul><li>Uniquely support of seamless switching between thinking mode (for complex logical reasoning, math, and coding) and non-thinking mode (for efficient, general-purpose dialogue) within single model, ensuring optimal performance across various scenarios.</li><li>Significantly enhancement in its reasoning capabilities, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.</li><li>Superior human preference alignment, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.</li><li>Expertise in agent capabilities, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.</li><li>Support of 100+ languages and dialects with strong capabilities for multilingual instruction following and translation.</li></ul><h2></h2>

Qwen3

qwen3_4b.safetensors

Z-Image is a powerful and highly efficient image generation model with 6B parameters. It is currently has three variants:<ul><li>🚀 Z-Image-Turbo – A distilled version of Z-Image that matches or exceeds leading competitors with only 8 NFEs (Number of Function Evaluations). It offers ⚡️sub-second inference latency⚡️ on enterprise-grade H800 GPUs and fits comfortably within 16G VRAM consumer devices. It excels in photorealistic image generation, bilingual text rendering (English &amp; Chinese), and robust instruction adherence.</li><li>🧱 Z-Image-Base (this model) – The non-distilled foundation model. By releasing this checkpoint, we aim to unlock the full potential for community-driven fine-tuning and custom development.</li><li>✍️ Z-Image-Edit – A variant fine-tuned on Z-Image specifically for image editing tasks. It supports creative image-to-image generation with impressive instruction-following capabilities, allowing for precise edits based on natural language prompts.</li></ul>Original ComfyUI Models: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image">https://huggingface.co/Comfy-Org/z_image</a>Original HF Repo: <a target="_blank" rel="ugc" href="https://huggingface.co/Tongyi-MAI/Z-Image">https://huggingface.co/Tongyi-MAI/Z-Image</a>

Z Image Base

Comfy Repack <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image">https://huggingface.co/Comfy-Org/z_image</a>

Files with SHA256: 7ca32dcf07dfe7692945d80fff86e3a74cb83c6206b9b223ac6836b939bb85d6