Zhipu AI Research https://www.zhipuai.cn/zh/research Research and technical blogs from Zhipu AI (智谱) http://www.rssboard.org/rss-specification RSS Creator zh-CN Thu, 24 Sep 2026 07:21:47 +0000 GLM-5.3-Flash：前沿智能进入普惠时代 https://www.zhipuai.cn/zh/research/163 基座模型 https://www.zhipuai.cn/zh/research/163 Wed, 26 Aug 2026 14:00:00 +0000 GLM-5.3：前沿编程能力与涌现的网络安全能力 https://www.zhipuai.cn/zh/research/162 基座模型 https://www.zhipuai.cn/zh/research/162 Fri, 14 Aug 2026 06:00:00 +0000 GLM-5.2上线并开源：专注Coding与长程任务 https://www.zhipuai.cn/zh/research/161 基座模型 https://www.zhipuai.cn/zh/research/161 Tue, 16 Jun 2026 16:00:00 +0000 下一代大模型推理网络架构：ZCube如何有效破解网络瓶颈？ https://www.zhipuai.cn/zh/research/160 ZCube：超大规模大模型推理的网络优化 https://www.zhipuai.cn/zh/research/160 Wed, 20 May 2026 19:51:00 +0000 Scaling Pain：超大规模Coding Agent推理实践 https://www.zhipuai.cn/zh/research/159 we share the lessons learned from this investigation, with the hope of helping the community better understand and overcome the Scaling Pain of Coding Agent inference. https://www.zhipuai.cn/zh/research/159 Wed, 29 Apr 2026 16:00:00 +0000 GLM-5.1开源：一个独立工作8小时的模型 https://www.zhipuai.cn/zh/research/157 our next-generation flagship model for agentic engineering https://www.zhipuai.cn/zh/research/157 Tue, 07 Apr 2026 16:00:00 +0000 GLM-5V-Turbo发布：多模态Coding基座模型 https://www.zhipuai.cn/zh/research/156 A multimodal coding foundation model for visual programming https://www.zhipuai.cn/zh/research/156 Wed, 01 Apr 2026 16:00:00 +0000 GLM-5-Turbo：龙虾增强的基座模型 https://www.zhipuai.cn/zh/research/155 A foundation model for OpenClaw https://www.zhipuai.cn/zh/research/155 Sun, 15 Mar 2026 16:00:00 +0000 GLM-5技术报告 https://www.zhipuai.cn/zh/research/153 to transition the paradigm of vibe coding to agentic engineering https://www.zhipuai.cn/zh/research/153 Sat, 21 Feb 2026 16:00:00 +0000 GLM-5开源：从代码到工程，Agentic Engineering时代最好的开源模型 https://www.zhipuai.cn/zh/research/154 targeting complex systems engineering and long-horizon agentic tasks https://www.zhipuai.cn/zh/research/154 Wed, 11 Feb 2026 16:00:00 +0000 GLM-OCR发布：性能SOTA，搞定复杂文档 https://www.zhipuai.cn/zh/research/150 a lightweight professional OCR model https://www.zhipuai.cn/zh/research/150 Mon, 02 Feb 2026 16:00:00 +0000 GLM-4.7-Flash开源、免费 https://www.zhipuai.cn/zh/research/148 Small but powerful. https://www.zhipuai.cn/zh/research/148 Mon, 19 Jan 2026 16:00:00 +0000 智谱联合华为开源国产芯片训练的多模态SOTA模型 https://www.zhipuai.cn/zh/research/158 the first open-source, industrial-grade discrete auto-regressive image generation model https://www.zhipuai.cn/zh/research/158 Tue, 13 Jan 2026 16:00:00 +0000 GLM-4.7：更强的 Coding https://www.zhipuai.cn/zh/research/143 GLM-4.7, your new coding partner, is coming with the following features: https://www.zhipuai.cn/zh/research/143 Sun, 21 Dec 2025 16:00:00 +0000 GLM-TTS：基于多奖励融合强化学习，实现工业级语音合成 https://www.zhipuai.cn/zh/research/147 Generating more expressive and emotional speech. https://www.zhipuai.cn/zh/research/147 Wed, 10 Dec 2025 16:00:00 +0000 GLM-ASR-Nano：面向真实世界的高鲁棒性语音识别 https://www.zhipuai.cn/zh/research/149 a robust, open-source speech recognition model with 1.5B parameters https://www.zhipuai.cn/zh/research/149 Tue, 09 Dec 2025 16:00:00 +0000 AutoGLM开源：每台手机，都可以成为AI手机 https://www.zhipuai.cn/zh/research/145 Unlocking the AI Phone for Everyone https://www.zhipuai.cn/zh/research/145 Mon, 08 Dec 2025 16:00:00 +0000 GLM-4.6V：支持原生工具调用的开源多模态模型 https://www.zhipuai.cn/zh/research/144 effectively bridges the gap between "visual perception" and "executable action" https://www.zhipuai.cn/zh/research/144 Sun, 07 Dec 2025 16:00:00 +0000