According to the Gaoling School of Artificial Intelligence at Renmin University of China: on 8 11, 2026, the second meeting of the Academic Committee of the Beijing Key Laboratory of Large Models and Intelligent Governance was held online. Laboratory director Wen Jirong reported on annual progress, presenting achievements including the Yulan hybrid architecture model, the diffusion language model LLaDA-MoE, and the Yulan Wanxiang social simulation platform. Among them, the Guantian nowcasting precipitation intelligent forecast system, developed in collaboration with the China Meteorological Administration, has been deployed at 9 provincial and municipal meteorological units. Committee members offered recommendations on directions such as the differences between diffusion and autoregressive models, novel architectures, and AI for Science.
The China Science and Technology Development Foundation, together with Shanghai AI Laboratory, has announced that the second "AI Research Accelerator Camp" free online training will be held from 2026 to 11-12. The training targets master's and doctoral students as well as research and technology industry professionals, comprising 21 sessions covering general AI research workflows, hands-on research tool practice, and application cases across six domains. Registration closes on 2026-11-22 at 24:00, and the program is entirely free with course replays available.
The 2026 State of AI in the Enterprise report released by Menlo Ventures on 3 11 2026 shows that 89% of US enterprises' open-source model spend flows to Chinese models: Kimi at 61%, DeepSeek at 15%, and Zhipu GLM at 11%. The article argues the drivers include the price gap (open-source average of about $0.13 per million tokens vs. about $0.70 for closed-source), shifts in token consumption structure driven by agents, and the control offered by open weights. It also mentions regulatory restrictions at the government procurement level.
According to Mysteel, citing the Meituan Tech WeChat account, Meituan open-sourced LongCat-2.5 on 10 11, 2026, a natively multimodal model for long-horizon agentic tasks, under the MIT license, allowing commercial inference deployment and further training without additional authorization. The model has 1.6 trillion total parameters, activates approximately 480 billion parameters per token, supports one million-token context, and has been optimized in architecture and deployment for domestic compute chips constrained in memory and bandwidth.
According to hot-topic coverage published by YeYuLingFeng Learning Network on 2026 11, 6, Tencent Hunyuan officially released the 11 large model on 4 Hy4, introducing for the first time visual understanding and computer operation (Computer Use) capabilities compared with the 8 Preview version. In a blind test of 213 tasks by 579 financial experts, Hy4 scored a composite of 3.04, higher than several open-source competitors; the hallucination rate fell from 12.95% to 7.2%. The company also acknowledged a significant gap versus top closed-source models, with the weight release date not yet announced.
According to Yicai, OpenAI announced on local time 11 4 the acquisition of developer-tools startup Pierre Computer Company, whose Code Storage, Trees, and Diffs tools have been integrated into its coding agent Codex; the deal amount was not disclosed. On the same day, Cat Wu (Catherine Wu), product lead for Anthropic's Claude Code and Cowork, left to join OpenAI. Combined with the simultaneously announced proprietary Git backend Pando, OpenAI is moving down the stack into development infrastructure such as code storage and version control, though it has not announced a direct rivalry with GitHub.
According to a People's Finance News report on 11月5, Huawei Cloud's MaaS platform will update and launch the openPangu-2.1-Pro flagship model on 11月15, with a focus on improving ArkTS code generation quality in HarmonyOS native app development; core products including the agent platform AgentArts, CodeArts, and OfficeAce will simultaneously integrate the model. The report also notes that earlier claims that Huawei Cloud had completely discontinued sales of the Pangu large model were a misreading.
The Qwen4 release window relayed by outside observers from 2026 11 2 to 8 has passed, and Alibaba's official written statement remains only the 9 22 remark that it is "in training," with no release date, model specs, pricing, or license announced. Verifiable facts in the article include the four model names from the 9 22 live screen projection, the context change merged in the Qwen CLI repository on 10 28, and the official poster data from 11 6. The article recommends moving Qwen4 off the release schedule and keeping only a watch position.
On 11/5, Baidu open-sourced its base model Wenxin 5.5 Flash Base (ERNIE 5.5 Flash Base). The model uses a MoE architecture with 2800 billion total parameters and 140 billion activated per token, was pretrained on 33.5 trillion tokens, and supports up to 1000K-token context during fine-tuning and inference. Baidu claims it achieves SOTA performance among same-size models on benchmarks covering knowledge, reasoning, math, code, multilingual, and long-context tasks.
SenseTime released the official version of SenseNova 6.8 Flash on 11月2日. According to SenseTime, the model is designed for complex tasks in real-world environments, supports cross-task and cross-session collaboration, and its delivery formats cover static web pages, dynamic content, and design deliverables. Evaluations claim it surpasses some leading international frontier models in agent collaboration, data analysis, and presentation generation scenarios. Reported by Shanghai Securities News.
The daily AI digest by blogger Yingchen Langdi AI Ke Hu aggregates multiple updates from late 10 2026: iFlytek open-sourced Spark X2.6, built on fully domestic compute and supporting million-token context; ModelBest, Xiaomi, Microsoft, and MemTensor released multimodal and memory-native models; funding updates for DeepSeek, OpenAI, and Anthropic; IDC data shows Huawei Ascend's share of China's AI chip market rose to 28%; and the API price war continues, with industry estimates putting inference costs down about 99.4% over two years. The content cites multiple media outlets and requires item-by-item verification.
Meituan's technical team official account announced the release of LongCat-2.5 on 11 2. According to the National Business Daily, the model targets long-horizon agent tasks, with 1.6 trillion total parameters and approximately 480 billion activated per token, strengthening native visual understanding and agentic coding capabilities on top of LongCat-2.0.