Z.ai opens GLM-5-Turbo to Coding Plan users

Z.ai Opens GLM-5-Turbo to Coding Plan Users

What Happened

Z.ai announced on Twitter: GLM-5-Turbo is now available to all Coding Plan subscribers, no longer limited to premium tiers.

GLM-5-Turbo is built on the open-source GLM-5 and is specifically optimized for Agent workflows: tool invocation, complex instruction breakdown, scheduled tasks, and long-chain execution. The pricing is as follows:

Why It Matters

• Product Direction Changed: GLM-5-Turbo targets autonomous, multi-step engineering tasks, which is currently the mainstream direction for AI development tools. The base version GLM-5 (744 billion parameters, 40 billion active parameters) already performs well on coding benchmarks: SWE-bench Verified 77.8%, Terminal-Bench 2.0 56.2%. The Turbo version has been specially accelerated for Agent scenarios and performs well on Z.ai’s internal ZClawBench (covering tasks like data analysis and automation).
• Pricing Below Mainstream: For similar workloads, GLM-5-Turbo is cheaper than Claude Haiku 4.5 and GPT-5.2. Z.ai has packaged it into a quarterly billing Coding Plan ($27—$216/quarter), which also includes tools like Vision Analysis and Web Search.
• Open Source with Commercial Optimization: GLM-5 remains fully open source under the MIT license, while Turbo is closed source and chargeable. Z.ai is betting that even if the base model is free, developers will still pay for engineering optimizations in Agent scenarios. Whether this “open-source foundation + proprietary enhancement” hybrid strategy can be sustained is being watched by other labs.

Impact Assessment

• Importance: High
• Category: Model Release, Developer Tools, AI Research

Conclusion: For teams looking to implement multi-step automation and Agent orchestration, this is a competitively priced early entry point. The most benefit will go to developers and corporate tech teams working on automation products and workflows; it is less relevant for trading participants and passive long-term holders.