glm-models

Call GLM models via a unified HTTP API for chat, thinking, streaming, and tool calls.

Updated Aug 23, 2026

Search Tags:#api #llm #streaming #thinking #tools #glm

Ready to run

npx skills add https://github.com/FeatherHunter/StudyNotes --skill glm-models

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: glm-models
Source: https://github.com/FeatherHunter/StudyNotes/tree/main/.qoder/skills/glm-models
Command: npx skills add https://github.com/FeatherHunter/StudyNotes --skill glm-models

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill encapsulates GLM model API calls (glm-4.7, glm-4.6, glm-4.5, etc.) in a single, reusable unit, enabling seamless integration of conversational AI, thinking chains, streaming outputs, and function calls without boilerplate.

Core Features & Use Cases

Unified GLM access: wrapper around several GLM variants for chat completions.
Thinking mode & streaming: supports Thinking Mode and streaming responses for interactive workflows.
Tool integration: supports external function calls to extend capabilities in AI prompts.
Use case examples: build AI assistants, code analysis tools, or complex reasoning tasks with prompt chaining.

Quick Start

Ask a GLM question using the provided commands to receive a response from the selected model.

Frequently Asked Questions about glm-models

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I call GLM models for chat completions with streaming responses?▼

You can call GLM models by sending an HTTP request with a specified model name, messages array, and optional parameters to an API endpoint that returns structured streaming responses for interactive workflows.

What is thinking mode in GLM and how does it handle complex reasoning?▼

Thinking mode in GLM enables advanced reasoning by processing prompts through structured thinking chains, allowing the model to break down complex tasks like code analysis or multi-step problem solving before returning a final response.

Can I use external function calls with GLM to extend AI assistant capabilities?▼

Yes, GLM supports tool integration by accepting optional tool configurations in the API request, enabling external function calls that extend the capabilities of AI assistants within your prompts.

Which GLM model variants are available for building AI assistants?▼

The API provides unified access to GLM-4.7, GLM-4.6, GLM-4.5 and related variants, allowing you to select the appropriate model name for your specific assistant, code analysis, or reasoning task.

Does GLM API require specific dependencies to enable streaming and tool calls?▼

No external dependencies are required to use the GLM API wrapper; it operates independently by exposing an HTTP endpoint that accepts model configurations and returns structured responses with streaming support.

What configurations should I set for GLM API requests to limit token output?▼

You can limit token output by including the max_tokens parameter in your GLM API request, alongside the required model name and messages, to control the length of the structured response returned.