GLM Agent Example

This example demonstrates how to use Agno-Go with GLM (智谱AI), China's leading domestic LLM platform.

Overview

GLM (Zhipu AI) is an advanced language model developed by Tsinghua University's Knowledge Engineering Group. It offers:

• Optimized for Chinese: Excellent performance on Chinese language tasks
• GLM-4: Main conversational model with 128K context
• GLM-4V: Vision-enabled multimodal capabilities
• GLM-3-Turbo: Fast and cost-effective variant

Prerequisites

1. Go 1.21+ installed
2. GLM API Key from https://open.bigmodel.cn/

Getting Your API Key

1. Visit https://open.bigmodel.cn/
2. Sign up or log in
3. Navigate to API Keys section
4. Create a new API key

The API key format is: {key_id}.{key_secret}

Installation

go get github.com/rexleimo/agno-go

Environment Setup

Create a .env file or export the environment variable:

export ZHIPUAI_API_KEY=your-key-id.your-key-secret

Basic Example

package main

import (
    "context"
    "fmt"
    "log"
    "os"

    "github.com/rexleimo/agno-go/pkg/agno/agent"
    "github.com/rexleimo/agno-go/pkg/agno/models/glm"
)

func main() {
    // Create GLM model
    model, err := glm.New("glm-4", glm.Config{
        APIKey:      os.Getenv("ZHIPUAI_API_KEY"),
        Temperature: 0.7,
        MaxTokens:   1024,
    })
    if err != nil {
        log.Fatalf("Failed to create GLM model: %v", err)
    }

    // Create agent
    agent, err := agent.New(agent.Config{
        Name:         "GLM Assistant",
        Model:        model,
        Instructions: "You are a helpful AI assistant.",
    })
    if err != nil {
        log.Fatalf("Failed to create agent: %v", err)
    }

    // Run agent
    output, err := agent.Run(context.Background(), "Hello! Tell me about yourself.")
    if err != nil {
        log.Fatalf("Agent run failed: %v", err)
    }

    fmt.Println(output.Content)
}

Example with Tools

package main

import (
    "context"
    "fmt"
    "log"
    "os"

    "github.com/rexleimo/agno-go/pkg/agno/agent"
    "github.com/rexleimo/agno-go/pkg/agno/models/glm"
    "github.com/rexleimo/agno-go/pkg/agno/tools/calculator"
    "github.com/rexleimo/agno-go/pkg/agno/tools/toolkit"
)

func main() {
    // Create GLM model
    model, err := glm.New("glm-4", glm.Config{
        APIKey:      os.Getenv("ZHIPUAI_API_KEY"),
        Temperature: 0.7,
        MaxTokens:   1024,
    })
    if err != nil {
        log.Fatal(err)
    }

    // Create agent with calculator tools
    agent, err := agent.New(agent.Config{
        Name:         "GLM Calculator Agent",
        Model:        model,
        Toolkits:     []toolkit.Toolkit{calculator.New()},
        Instructions: "You are a helpful AI assistant that can perform calculations.",
    })
    if err != nil {
        log.Fatal(err)
    }

    // Test calculation
    output, err := agent.Run(context.Background(), "What is 123 * 456?")
    if err != nil {
        log.Fatal(err)
    }

    fmt.Printf("Result: %s\n", output.Content)
}

Chinese Language Example

GLM excels at Chinese language tasks:

package main

import (
    "context"
    "fmt"
    "log"
    "os"

    "github.com/rexleimo/agno-go/pkg/agno/agent"
    "github.com/rexleimo/agno-go/pkg/agno/models/glm"
)

func main() {
    model, err := glm.New("glm-4", glm.Config{
        APIKey:      os.Getenv("ZHIPUAI_API_KEY"),
        Temperature: 0.7,
    })
    if err != nil {
        log.Fatal(err)
    }

    agent, err := agent.New(agent.Config{
        Name:         "中文助手",
        Model:        model,
        Instructions: "你是一个有用的中文AI助手。",
    })
    if err != nil {
        log.Fatal(err)
    }

    // Ask in Chinese
    output, err := agent.Run(context.Background(), "请用中文介绍一下人工智能的发展历史。")
    if err != nil {
        log.Fatal(err)
    }

    fmt.Println(output.Content)
}

Running the Example

1. Clone the repository:

git clone https://github.com/rexleimo/agno-go.git
cd agno-Go

2. Set your API key:

export ZHIPUAI_API_KEY=your-key-id.your-key-secret

3. Run the example:

go run cmd/examples/glm_agent/main.go

Configuration Options

glm.Config{
    APIKey:      string  // Required: {key_id}.{key_secret} format
    BaseURL:     string  // Optional: Custom API endpoint
    Temperature: float64 // Optional: 0.0-1.0 (default: 0.7)
    MaxTokens:   int     // Optional: Max response tokens
    TopP:        float64 // Optional: Top-p sampling parameter
    DoSample:    bool    // Optional: Enable sampling
}

Authentication

GLM uses JWT (JSON Web Token) authentication:

• API key is split into key_id and key_secret
• JWT token is generated using HMAC-SHA256 signing
• Token is valid for 7 days
• Automatically handled by the SDK

Supported Models

Model	Context	Best For
glm-4	128K	General conversation, Chinese language
glm-4v	128K	Vision tasks, multimodal
glm-3-turbo	128K	Fast responses, cost-effective

Common Issues

Invalid API Key Format

Problem: API key must be in format {key_id}.{key_secret}

Solution: Ensure your API key contains a dot (.) separator between key_id and key_secret.

Authentication Failed

Problem: GLM API error: Invalid API key

Solution:

• Verify your API key is correct
• Check if the API key is active at https://open.bigmodel.cn/
• Ensure no extra spaces in the environment variable

Rate Limiting

Problem: GLM API error: Rate limit exceeded

Solution:

• Implement retry logic with exponential backoff
• Reduce request frequency
• Upgrade your API plan if needed

Next Steps

• Learn about Models for more LLM options
• Add more Tools to enhance capabilities
• Build Teams with multiple agents
• Explore Workflows for complex processes

Related Examples

• Simple Agent - OpenAI example
• Claude Agent - Anthropic example
• Team Demo - Multi-agent collaboration

Resources

• GLM Official Website
• GLM API Documentation
• Agno-Go Repository