What it does
The Z.ai Open Platform lets developers call the company's models from their own code. The quick start walks through four steps: register or log in, create a Standard API key, choose a model, and make a call. Billing can be topped up from the platform's billing page. Two models are highlighted. GLM-5.3 is described as the latest flagship model, with advances in software engineering and agent capabilities. GLM-5.3-Flash is described as a multimodal coding model that specializes in visual programming. You can call the API over standard HTTP, with the official Z.ai Python or Java SDKs, or with the OpenAI Python, Node.js or Java SDKs by pointing them at Z.ai's base URL. The documentation navigation also covers thinking mode, streaming, function calling, context caching, structured output, web search, and agents for slides and posters (beta), translation and video effect templates. The page also promotes a GLM Coding Plan, a monthly subscription for coding and office tasks that works with more than 20 agent tools, from $18 a month. Per-token API prices are on a separate pricing page that is not part of the text read here.
Capabilities
- HTTP API plus official Python and Java SDKs
- OpenAI SDK compatibility for migrating existing code
- GLM-5.3 and GLM-5.3-Flash models
- Thinking mode, streaming, function calling, context caching and structured output
- Web search tool and agents for slides and posters, translation and video effects
- GLM Coding Plan subscription for coding agents
When to use it
- You want to add GLM models to an application through an API
- You already use the OpenAI SDK and want to try another provider with few code changes
- You want a monthly plan for coding and office tasks across agent tools
