GLM-5.1(于 2026 年 4 月发布),专为长时程自主任务打造。不同于针对短时交互优化的传统模型,GLM-5.1 在保持目标对齐、降低策略漂移、并在长时间跨度内交付生产级结果方面表现出色—在单个复杂任务上可连续自主运行长达 8 小时。这标志着智能体工程的一次重大飞跃,将评估从单轮智能转向真实世界的持续执行。 上下文:200k
GLM-5 Turbo 是 Z.ai 推出的一款新模型,专为快速推理和在 OpenClaw 等由智能体驱动的环境中的强大性能而设计。GLM-5 是 Z.ai 的旗舰开源基础模型,专为复杂系统设计与长时程智能体工作流而打造。
面向资深开发者,它在大规模编程任务上提供生产级性能,可与领先的闭源模型相媲美。
凭借先进的智能体规划、深度后端推理与迭代式自我纠错,GLM-5 不止于代码生成,能够实现全系统构建与自主执行。 上下文:200,000
Zhipu 最新旗舰模型 GLM-4.6 发布:总参数 355B,活跃参数 32B。整体核心能力超过 GLM-4.5。编程:与 Claude Sonnet 4 对齐,国内最佳。上下文:扩展至 200K(原为 128K)。推理:增强,支持 Tool 调用。搜索:优化 Tool 与 agent 框架。写作:更贴合人类偏好、写作风格与角色扮演。多语言:翻译效果增强。上下文:200K
GLM-4.7 is Z.AI’s latest flagship model, featuring upgrades in two key areas: enhanced programming capabilities and more stable multi-step reasoning/execution. It demonstrates significant improvements in executing complex agent tasks while delivering more natural conversational experiences and superior front-end aesthetics.High-performance, strong Inference, extremely fast response model, optimized for scenarios requiring ultra-fast Inference speed and powerful logical capabilities, providing millisecond-level response experience. GLM-4.5-Flash is an artificial intelligence model provided by ZhipuAI. Lightweight, high-performance, ultra-fast response model, perfectly combining the cost advantages of Air and the speed advantages of X, an ideal choice for balancing performance and efficiency. GLM-4.5-Air is an artificial intelligence model provided by ZhipuAI. GLM-4.5 is an artificial intelligence model provided by ZhipuAI.