ModelsMulti-sourceSelected73°

Qwen-Image 2.1 gets day-0 serving in SGLang-Diffusion

Why we picked it

A consumer RTX 4090 24GB runs Qwen-Image 2.1 at native precision in 18.7 seconds per 1024×1024 image, giving self-hosters a cheaper route than renting cloud image-generation APIs.

Alibaba's Qwen team announced that SGLang-Diffusion now offers day-0 support for Qwen-Image 2.

Body

1, handling text-to-image generation, multi-image editing and transparent RGBA output from one checkpoint.On a single RTX 4090 24GB with CPU offload, the model produces a 1024×1024 image in 18.

7 seconds and completes editing in 21.7 seconds, with 22.

7 GiB peak GPU memory during requests.On an RTX PRO 6000 96GB, generation takes 8.

0 seconds and editing 9.6 seconds.

Serving runs at native precision with no quantization, using 40 denoising steps per request, and supports TP/SP, LoRA and OpenAI-compatible APIs.

Original · 阿里通义 Qwen

Thanks @sgl_project for the day-0 support! 🙌 SGLang-Diffusion now serves Qwen-Image-2.1: text-to-image generation, multi-image editing, and transparent RGBA output. Try it out! 🎨 SGLang @sgl_project Day-0 support for @Alibaba_Qwen ’s Qwen-Image 2.1 is here in SGLang-Diffusion! 🖥️ Native precision on a single RTX 4090 24GB with CPU offload - 1024×1024 generation in 18.7s and image editing in 21.7s with 22.7 GiB peak GPU memory during requests. - On an RTX PRO 6000 96GB: 8.0s generation and 9.6s editing. 🎨 Text-to-image, multi-image editing, and transparent RGBA output—all with one checkpoint. ⚡ Native inference with TP/SP, LoRA, and OpenAI-compatible APIs. 40 denoising steps, one image per request, warmed HTTP latency including PNG output. No quantization. Cookbook and GPU-specific commands below 👇 🔗 View Quoted Tweet📊 11 ⚡

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Qwen-Image 2.1, SGLang, image generation, image editing.

Source attribution: curated from 阿里通义 Qwen (@Alibaba_Qwen). Cross-source corroboration: 11 independent sources have covered this development (including shao__meng, lmarena.ai, Decoder), confirming industry-wide signal strength.