--- format: "aidr-story-markdown/v1" id: "45a7a18bfbb87276463c0091204f36e3f0c3ff5199d2ad0e391803401f10d11e" canonical_url: "https://aidr.today/45a7a18b?lang=en" title: "Alibaba Releases 8 Step Image Generation for Open Weight Qwen 2.1" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-09T14:37:29.000Z" category: "Open Source" topics: ["open-source","inference","devtools"] source_urls: ["https://huggingnews.com/ai/alibaba-releases-8-step-image-generation-for-open-weight-qwen-21-bff365b7","https://x.com/Alibaba_Qwen/status/2108549075218120949","https://marketbrief.now/ai/alibaba-releases-8-step-image-generation-for-open-weight-qwen-21-bff365b7","https://www.marktechpost.com/2026/10/09/alibaba-qwen-releases-qwen-image-2-1-turbo-an-8-step-7b-image-model/"] summary: "Alibaba released the Qwen-Image-2.1-Turbo model as an open weights release, which uses a 7B visual generation architecture to create and edit 2K images in 8 steps. The company also launched Pro and Turbo APIs, while the foundational Qwen-Image-2.1 model supports native RGBA layer generation and up to 10 reference images." --- # Alibaba Releases 8 Step Image Generation for Open Weight Qwen 2\.1 > [Open the canonical story]() **Published:** 2026-10-09T14:37:29.000Z **Category:** Open Source **Topics:** open\-source, inference, devtools ## Summary Alibaba released the Qwen\-Image\-2\.1\-Turbo model as an open weights release, which uses a 7B visual generation architecture to create and edit 2K images in 8 steps\. The company also launched Pro and Turbo APIs, while the foundational Qwen\-Image\-2\.1 model supports native RGBA layer generation and up to 10 reference images\. ## Sources - [Story source]() - [Story source]() - [Story source]() - [Story source]()