- Zhipu releases the GLM-5-Turbo foundation model optimized for OpenClaw, aiming to solve the challenges of general large models stalling in complex, long-chain tasks.
- Zhipu has launched Claw subscription plans and a Claw security management system for enterprise scenarios.
Chinese AI startup Zhipu has launched GLM-5-Turbo, a new foundation model deeply optimized for OpenClaw scenarios.
The model aims to fundamentally solve the core challenges where general large models easily stall during long-chain tasks when entering real, complex agent scenarios, according to a statement by Zhipu on Monday.
The release of the core foundation model marks an increasingly fierce and white-hot market competition among major Chinese tech giants in the desktop AI sector.
Shortly before this, industry giants including ByteDance, Tencent, and Alibaba successively launched their own desktop AI agents or related cloud platform products.
Even if general models perform well in conversational capabilities, they are still prone to instruction-following deviations, unstable tool calling, and mid-way stalling in long tasks within OpenClaw scenarios, Zhipu said.
To fundamentally resolve these challenges, deep optimization must be conducted at the foundation model layer — which is the starting point for GLM-5-Turbo, the company noted.
Starting from the underlying training phase, GLM-5-Turbo has undergone deep, targeted optimization for tool calling, complex instruction following, and scheduled continuous tasks.
In its in-house developed benchmark ZClawBench, the model overall leads several mainstream domestic competitors across multiple key tasks, Zhipu said.
To address the high token consumption caused by long-chain tasks, Zhipu has introduced exclusive, flexibly subscribed plans for both enterprise and individual users.
In addition, its new enterprise-grade security management system supports centralized control of agent tasks, thereby ensuring the compliance and security of sensitive business data.
Currently, AI agents like OpenClaw are accelerating their entry into organizations, allowing every enterprise to have exclusive, 24/7 online digital employees.
Large model technology is undergoing a highly critical industry leap, transforming from a mere efficiency-boosting tool into a core enterprise workforce.
