OpenAI Agents Go SDK v0.1.0

First public release of the OpenAI Agents Go SDK - a lightweight yet powerful framework for building multi-agent workflows in Go.

Overview

This is a Go port of the OpenAI Agents Python SDK, designed to be as close as possible to the original implementation in both behavior and API.

Features

Core Functionality

• Multi-Agent Workflows: Build sophisticated agent systems with handoffs and coordination
• Provider-Agnostic Design: Works with OpenAI API and other LLM providers
• Function Tools: Full support for function calling and custom tools
• Guardrails: Configurable safety checks for input and output validation
• Session Management: Persistent memory across multiple runs

Advanced Capabilities

• Model Context Protocol (MCP): Integration with hosted and local MCP servers
• Voice Support: Both static and streaming voice responses
• Built-in Tools: Code interpreter, computer use, file search, web search
• Workflow Runner: Experimental orchestration layer for complex multi-agent workflows

Examples & Documentation

Comprehensive examples covering:

• Basic usage (hello world, streaming, templates, tools)
• Agent patterns (routing, guardrails, parallelization)
• Real-world scenarios (customer service, financial research, research bot)
• MCP integration (filesystem, git, prompts)
• Model providers (custom providers, LiteLLM)

Installation

go get github.com/nlpodyssey/[email protected]

Quick Start

package main

import (
    "context"
    "fmt"
    "github.com/nlpodyssey/openai-agents-go/agents"
)

func main() {
    agent := agents.New("Assistant").
        WithInstructions("You are a helpful assistant").
        WithModel("gpt-4o")

    result, err := agents.Run(context.Background(), agent, "Write a haiku about recursion in programming.")
    if err != nil {
        panic(err)
    }

    fmt.Println(result.FinalOutput)
}

Requirements

• Go 1.24.3 or higher
• OpenAI API key (set OPENAI_API_KEY environment variable)

What's Next

This is a v0.x release, indicating that the API may still evolve based on community feedback and the upstream Python SDK development. We welcome contributions, bug reports, and feature requests!

Acknowledgments

Special thanks to:

• The OpenAI team for the original Python SDK
• All contributors who helped shape this first release
• The Anthropic team for the Model Context Protocol

For detailed documentation and examples, visit the repository.