Yazılar

China’s Moonshot AI Launches Open-Source Model to Regain Market Share

/in /tarafından

Chinese AI startup Moonshot AI unveiled its new open-source model, Kimi K2, on Friday, aiming to regain traction in the highly competitive domestic AI market. The model is designed with advanced coding skills and excels in general agent tasks and tool integration, enabling it to handle complex workflows more efficiently, the company said in a statement.

Moonshot claims Kimi K2 surpasses several mainstream open-source AI models, including DeepSeek’s V3, and competes closely with top U.S. models like Anthropic’s in certain coding-related functions. This release aligns with a growing trend among Chinese AI firms to open-source their models, contrasting with many U.S. tech giants, such as OpenAI and Google, which keep their most advanced AI technologies proprietary. However, some American companies like Meta have also embraced open-source AI models.

Open-sourcing helps companies build stronger developer communities, showcase technological prowess, and extend global influence—a strategic move by China to counter U.S. efforts to restrict its tech progress. Other Chinese tech giants that have open-sourced models include DeepSeek, Alibaba, Tencent, and Baidu.

Founded in 2023 by Tsinghua University graduate Yang Zhilin, Moonshot has become a key player in China’s AI scene, supported by major investors like Alibaba. The startup gained significant attention in 2024 for its long-text analysis and AI search capabilities but has seen its market position weaken after DeepSeek launched competitive low-cost models early this year.

According to the Chinese AI tracking site aicpb.com, Moonshot’s Kimi app was the third most-used AI product by monthly active users in August last year but slipped to seventh place by June 2025.

Huawei’s AI Lab Denies Copying Alibaba’s Qwen Model Amid Copyright Claims

/in /tarafından

Huawei’s AI research division, Noah Ark Lab, has denied allegations that its Pangu Pro Moe (Mixture of Experts) large language model copied from Alibaba’s Qwen 2.5 14B model. The lab insisted on Saturday that Pangu Pro was independently developed and trained, refuting claims made in a report by an entity named HonestAGI.

HonestAGI published a paper on GitHub claiming “extraordinary correlation” between Huawei’s Pangu Pro Moe and Alibaba’s Qwen model, suggesting that Huawei’s model might have been “upcycled” rather than trained from scratch. The report also raised concerns about potential copyright violations and false claims regarding Huawei’s investment in the model’s training.

In response, Noah Ark Lab stated that their model is not based on incremental training from other manufacturers’ models but instead includes key innovations in architecture and technical features. They highlighted that Pangu Pro is the first large-scale model built entirely on Huawei’s Ascend chips and confirmed adherence to open-source licensing rules for any third-party code used—though they did not specify which open-source models influenced their work.

Alibaba has yet to comment on the allegations, and the identity of HonestAGI remains unknown. The controversy comes amid rising competition in China’s AI sector, which has been accelerated by the release of open-source models like DeepSeek’s R1 and Alibaba’s Qwen family, designed for consumer and chatbot applications. In contrast, Huawei’s Pangu models are primarily applied in government, finance, and manufacturing sectors.

Rednote Joins Wave of Chinese Firms Releasing Open-Source AI Models

/in /tarafından

Chinese social media platform Rednote (Xiaohongshu) has released an open-source large language model named dots.llm1, joining a growing number of Chinese tech companies making AI models publicly available. This open-source move contrasts with many U.S. tech giants like OpenAI and Google, which keep their most advanced AI models proprietary, although some American firms such as Meta have also embraced open-source AI.

The release aims to showcase China’s technological prowess, foster developer communities, and extend global influence amid U.S. export restrictions targeting China’s advanced semiconductor industry.

According to Rednote’s technical paper published last Friday on Hugging Face, dots.llm1 performs comparably on coding tasks to Alibaba’s Qwen 2.5 model but is less advanced than models like DeepSeek-V3.

Rednote, known for its Instagram-style platform where users share photos, videos, and text, ramped up AI development after OpenAI’s ChatGPT debut in late 2022. Recently, it launched Diandian, an AI-powered search app for its main platform.

Other Chinese companies following this open-source path include Alibaba, which introduced the upgraded Qwen 3 model in April, and startup DeepSeek, whose low-cost R1 model has made waves globally for its competitive performance despite lower development costs.