Zhipu AI releases GLM-5.3 model focused on coding and cybersecurity
Zhipu AI announced the release of GLM-5.3, a model focused on programming and cybersecurity capabilities, post-trained on a 743B base model. The model has also been loaded on the Qoder platform, with reports indicating it surpasses GLM-5.2 on multiple benchmarks.
Detected & updated continuously · Source: Nebula
Story subjects
Sources
@PANewsCN
智普发布 GLM-5.3：主打编程和网络安全能力，在多项基准测试中超越 GLM-5.2 https://t.co/F9VjxRd2Cs 宣布推出 GLM-5.3 模型，主打编程和网络安全能力，GLM-5.3 通过在 743B 基础模型上进行后训练，实现了顶级编程和智能体能力，并在网络安全领域取得重大进展，为开源模型树立新标准。 https://t.co/C3Nfydqb6F
@NFT_Chen
🚨 炸了！GLM-5.3 已经偷偷上线 Qoder！ 刚实测： 调用 GLM-5.3 返回的是「必须开思考」错误码， 而不是「模型不存在」——后端已加载！ Qoder 模型列表直接出现： ✅ 智谱开源旗舰 ✅ 代码能力对标国际顶尖 ✅ 安全审查 + 漏洞推理 ✅ 200K 上下文 + max 思考 ✅ 仅 0.6x Credit 当年 GLM-5.2 https://t.co/8PxGMQerx6
@notjazii
chinese ai labs are catching up fast last month moonshot dropped kimi k3, the biggest open weight model yet and it came really close to the top frontier models now rumors are that zAI new model GLM 5.3 is coming today or later this week but chinese labs aren’t the only ones https://t.co/r1bv5bliif