GLM
Deal eligibility requirements
- Tier 1: monthly spend below $1,000.
- Tier 2: monthly spend of $1,000–$10,000.
- Tier 3: monthly spend above $10,000, or a 3-month prepayment.
- Tiers 2 and 3 require agreed prepaid or committed-volume terms. Discounts are already reflected in the published tier rates.
- A NachoCard is required to redeem this discount.
What is GLM?
GLM is Z.ai’s family of AI language models for software development, reasoning, and complex workflows. GLM-5.3 and GLM-5.2 help developers and engineering teams work with extensive project information, solve technical problems, and carry out tasks that involve multiple connected steps.
GLM-5.3
GLM-5.3 builds on the same base model as GLM-5.2, with additional training focused on complex coding, extended agent workflows, and cybersecurity reasoning. It is designed for demanding engineering work, including investigating problems, implementing changes, and evaluating results.
With a one-million-token context window, GLM-5.3 can work with substantial amounts of code, documentation, and task history. Its low, high, and maximum reasoning settings allow teams to adjust the depth of analysis to the task. Z.ai’s evaluations also show improved coding performance and output-token efficiency compared with GLM-5.2.
GLM-5.2
GLM-5.2 is designed for software projects that require sustained reasoning across large amounts of information. Its one-million-token context window supports working with codebases, technical documentation, architectural requirements, and previous decisions within an extended workflow.
The model can assist with code generation, debugging, refactoring, and application development. When connected to appropriate tools, it can support work from planning and implementation through testing and verification. Configurable reasoning effort and downloadable model weights give teams flexibility in how they integrate and deploy the model.
Key benefits of using GLM?
GLM-5.3
- Tackle demanding engineering tasks: Support complex programming and troubleshooting with a model trained on realistic development environments and tasks involving multiple dependencies.
- Carry work across multiple stages: Use stronger agent capabilities to connect investigation, implementation, and verification within longer workflows, while keeping the broader objective in context.
- Make reasoning effort fit the task: Choose low, high, or maximum effort depending on the complexity of the problem and the balance you need between depth and responsiveness.
- Support code-security analysis: Assist security professionals with identifying and investigating potential vulnerabilities in code, adding another analytical tool to security review workflows.
- Improve token efficiency: Z.ai reports stronger coding results with fewer output tokens than GLM-5.2 in its internal evaluations, which may help teams use their available token budgets more effectively.
GLM-5.2
- Keep more project information together: Work with large collections of code, requirements, and documentation so the model can consider relationships across different parts of a project.
- Manage changes across multiple files: Support refactoring, API migrations, and other development tasks that require coordinated updates and a clear understanding of dependencies.
- Maintain engineering consistency: Provide coding conventions, architectural constraints, and testing requirements to guide the model throughout an extended development task.
- Support broader development workflows: Apply coding and reasoning capabilities to web applications, mobile projects, debugging, and technical research.
- Gain deployment flexibility: Access model weights under the MIT license, allowing teams to explore customization and deployment on infrastructure suited to their requirements
GLM Features
- One-million-token context: Both models support extensive text inputs and long conversation histories.
- Up to 128K output tokens: Generate substantial responses for detailed explanations and larger coding tasks.
- Text-based interaction: Process natural-language instructions, source code, and technical documentation.
- Configurable reasoning: Adjust reasoning effort to match task requirements; GLM-5.3 offers low, high, and maximum settings.
- Function calling: Connect the models to external tools and services within your application.
- Structured outputs: Generate formats such as JSON for use in automated workflows.
- Streaming responses: Display output progressively as it is generated.
- Context caching: Reuse supported context across requests in longer interactions.
- API integration: Incorporate GLM capabilities into coding assistants, internal tools, and business applications.
GLM Pricing
USAGE PRICING — USD PER 1 MILLION TOKENS
Each line lists input / cached input / output.
GLM-5.3
Provider reference: $1.40 / $0.26 / $4.40
Tier 1 — 15% off: $1.19 / $0.221 / $3.74
Tier 2 — 20% off: $1.12 / $0.208 / $3.52
Tier 3 — 25% off: $1.05 / $0.195 / $3.30
GLM-5.2
Provider reference: $1.40 / $0.28 / $4.40
Tier 1 — 15% off: $1.19 / $0.238 / $3.74
Tier 2 — 25% off: $1.05 / $0.21 / $3.30
Tier 3 — 35% off: $0.91 / $0.182 / $2.86
TIER REQUIREMENTS
Tier 1: monthly spend below $1,000; self-serve top-up.
Tier 2: monthly spend of $1,000–$10,000; prepay / committed volume with weekly invoicing.
Tier 3: monthly spend above $10,000, or a 3-month prepayment; prepay / committed volume with weekly invoicing.
Provider reference prices and rates are from September 2026 and move with provider price changes. These are usage rates, not monthly subscription fees. Tier discounts are already included.