Zhipu GLM-4.5-Air
In order to enable every developer and user to easily experience the capabilities of cutting-edge large models,Zhipu has made the GLM-4.5-Air model available for free to Cherry Studio usersAs an efficient foundation model built specifically for agent applications, GLM-4.5-Air strikes an excellent balance between performance and cost, making it the ideal choice for building intelligent applications.
🚀 What is GLM-4.5-Air?
GLM-4.5-Air is Zhipu's latest high-performance language model, adopting an advanced Mixture-of-Experts (MoE) architecture, while maintaining outstanding reasoning capabilities, it significantly reduces compute resource consumption.
Total parameters: 106 billion
Active parameters: 12 billion
Through a streamlined design, GLM-4.5-Air achieves higher inference efficiency, making it suitable for deployment in resource-constrained environments while still being capable of handling complex tasks.
📚 Unified training pipeline, laying a solid intelligent foundation
GLM-4.5-Air shares the same training pipeline as the flagship series, ensuring a solid foundation of general capabilities:
Large-scale pretraining：up to 150 trillion tokens of general corpus it completed training on, building broad knowledge understanding capabilities;
Domain-specific optimization：strengthened training on key tasks such as code generation, logical reasoning, and agent interaction;
Long-context support：context length extended to 128K tokens, capable of handling long documents, complex conversations, or large code projects;
Reinforcement learning enhancement：through RL, the model's decision-making abilities in reasoning, planning, tool use, etc. are optimized.
This training system gives GLM-4.5-Air excellent generalization and task adaptability.
⚙️ Core capabilities optimized for agents
GLM-4.5-Air has been deeply adapted for agent application scenarios and has the following practical capabilities:
✅ Tool calling support：can call external tools through standardized interfaces to automate tasks ✅ Web browsing and information extraction：can work with browser plugins to understand and interact with dynamic content ✅ Software engineering assistance：supports requirement analysis, code generation, bug identification and fixing ✅ Front-end development support：has a good understanding of and generation capability for front-end technologies such as HTML, CSS, and JavaScript
The model can be flexibly integrated into Claude Code, Roo Code and other code-agent frameworks, and can also be used as the core engine of any custom Agent.
💡 Intelligent "thinking mode", flexibly responds to various requests
GLM-4.5-Air supports hybrid reasoning mode, users can use thinking.type parameter to control whether deep thinking is enabled:
enabled: enable thinking, suitable for complex tasks requiring step-by-step reasoning or planningdisabled: disable thinking, used for simple queries or instant responsesdefault is set to dynamic thinking mode, the model automatically decides whether in-depth analysis is needed
Simple tasks(recommended to turn off thinking)
- Query the founding time of "Zhipu AI" - Translate "I love you" into Chinese
Medium tasks(recommended to enable thinking)
- Compare the pros and cons of taking a plane vs. high-speed rail from Beijing to Shanghai - Explain why Jupiter has so many moons
Complex tasks(strongly recommended to enable thinking)
- Explain how experts collaborate in an MoE model - Analyze whether to buy an ETF based on market information
🌟 Efficient and low-cost, easier to deploy
GLM-4.5-Air strikes an excellent balance between performance and cost, especially suitable for real-world business deployment:
⚡ Generation speed exceeds 100 tokens/second, fast response, supports low-latency interaction
💰 API cost is extremely low: input only 0.8 yuan/million tokens, output 2 yuan/million tokens
🖥️ Few active parameters, low compute requirements, easy to run locally or in the cloud at high concurrency
Truly delivers an AI service experience of "high performance, low barrier to entry".
🧠 Focus on practical capability: intelligent code generation
GLM-4.5-Air performs stably in code generation and supports:
covering Python, JavaScript, Java and other mainstream languages
Generating according to natural language instructions code that is clearly structured and highly maintainable code
reduces templated output, closer to the needs of real development scenarios
Suitable for high-frequency development tasks such as rapid prototyping, autocomplete, and bug fixing.
Try it for free now GLM-4.5-Air, and start your agent development journey! Whether you want to build an automated assistant, a coding companion, or explore next-generation AI applications, GLM-4.5-Air will be your efficient and reliable AI engine.
📘 Connect now and unleash creativity!
Get help and submit feedback
If you encounter any questions, bugs, or have suggestions for feature improvements during configuration or use, please refer to Feedback and Suggestions for the official channels provided.
Last updated
Was this helpful?