Gate News Report, March 9 — According to LatePost, Alibaba announced management adjustments for the Qwen team today. Alibaba Cloud CTO and Tongyi Laboratory Director Zhou Jingren will oversee the Qwen Model No. 1; Liu Dayi Heng, responsible for pre-training, will also take over post-training and the Coding team, reporting to Zhou Jingren. This adjustment follows the sudden departure of former Qwen leader Lin Junyang last week. Alibaba stated that they will continue to adhere to their open-source strategy, and there will be no significant changes in the collaboration model for core modules such as pre-training and post-training.

PANews March 5 News, Google DeepMind Developer Experience Lead Omar Sanseviero posted, saying "Qwen friends…please reach out," inviting Tongyi Qianwen team members to join and participate in the open-source model ecosystem. The background of the news is that the recent resignation of core Qianwen members such as Lin Junyang has attracted attention.

PANews February 27 News, in the future, all new AI glasses launched by Alibaba will be uniformly branded as "Qwen AI Glasses" (Qwen Glasses) for the global market. The already launched Quark AI Glasses will be updated in functionality to stay in sync with Qwen AI Glasses and will enjoy Qwen AI services.

PANews February 16 News, according to Jinshi reports, sources revealed that Alibaba will open source the new generation Qwen 3.5 large model on New Year's Eve tonight. The model has achieved a comprehensive innovation in model architecture and is expected to mark a new milestone for domestic models.

Golden Finance reported that on December 22, Alibaba Tongyi Qianwen announced the launch of the image generation model Qwen-Image-Layered. The new model uses a self-developed architecture, which can "decompose" an image into multiple layers, each of which can be manipulated independently without affecting other content.

BlockBeats News, December 10 — After investing hundreds of billions of dollars to build the most expensive team in tech history for months, Meta CEO Zuckerberg is now deeply involved in daily R&D and is pushing the company's strategy toward directly monetizable AI models. According to sources, a new model codenamed "Avocado" is expected to be released in Spring 2026 and may be launched in a closed-source manner (i.e., strictly controlled by Meta and sold for access). This move marks a significant deviation from Meta's long-standing advocacy of an open-source approach. Zuckerberg is dedicating a large amount of time to a core team called TBD Lab, which, during the training of Avocado, even integrated third-party models including Google Gemma, OpenAIgpt-oss, and Alibaba Qwen. Meanwhile, M

PANews, November 25th - According to Shanghai Securities News, the Artificial Intelligence Singapore (AISG) has abandoned the Meta Llama model in its latest Southeast Asian language model project, opting instead for Alibaba's Tongyi Qwen architecture. As of now, the Qwen series has surpassed 600 million downloads globally. It is reported that the "Qwen-SEA-LION-v4" released by AISG on November 25th ranks first in Southeast Asian language capabilities. Previously, open source models represented by Meta's Llama series performed poorly in handling regional languages such as Indonesian, Thai, and Malay, severely limiting the development efficiency and performance of localized AI applications.

Alibaba plans to restructure its mobile AI application in the coming months, with the updated app being closer to ChatGPT and named "Qwen." The plan includes AI agent functionalities and shopping support, aiming to develop Qwen into a comprehensive AI agent and to launch an overseas version to expand the market.

Golden Finance reported that Jay A, the founder of nof1.ai, announced the end of Season 1 of Alpha Arena, with Qwen 3 MAX as the winner. Despite model biases, Season 1.5 will undergo improvements. Additionally, Julian Togelius from nofi.ai mentioned that the next season will optimize the AI Benchmark and launch new projects.

Odaily News nof1 founder Jay A announced on X that the Alpha Arena Season 1 competition officially ended this morning, with Qwen 3 MAX surpassing Deepseek Chat V3.1 at the last moment to become the AI model with the highest final profit.