Alibaba's Qwen team has released Qwen-Image-3.0, the third generation of their image generation foundation model. This new model boasts enhanced capabilities, including support for up to 4.5k token inputs, precise rendering of small text down to 10px, and native rendering for 12 languages. The announcement was made by Alibaba Cloud, with mentions of potential future models like GPT-6 and events like WAIC 2026. AI
IMPACT Sets a new benchmark for multilingual text rendering in AI image generation, potentially influencing future model development.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
