Alibaba has launched its latest foundational large language model, Qwen3.8, boasting a total of 2.4 trillion parameters. This new model shows significant improvements in coding and professional office tasks, positioning it as a strong competitor, currently ranking just below Anthropic's Claude series on the Arena leaderboard. Qwen3.8 is now accessible via API on the Qianwen AI platform and integrated into Alibaba's new Agent product, "Qianwen Office." The company also plans to open-source Qwen3.8-Max and Qwen3.8-27B next week. AI
IMPACT Sets a new benchmark for large parameter models, potentially driving competition in enterprise AI solutions.
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 1 sources. How we write summaries →
