OpenGithubModelsApi.jl

OpenGithubModelsApi.jl is a Julia client library for interacting with GitHub's AI Models API (https://models.github.ai). This package was automatically generated using OpenAPI.jl, providing a type-safe and validated interface to GitHub's machine learning models.

Features

• 🚀 Type-safe API client for GitHub's AI models
• 🔒 Secure authentication with GitHub tokens
• 💬 Chat completion functionality
• 🏢 Organization-specific model access
• 📚 Model catalog listing
• ⚙️ Comprehensive parameter validation
• 📦 Well-structured Julia package with proper documentation

Create your personal GitHub access tokens

Github Docs

Installation

To install the package, use Julia's package manager:

using Pkg
Pkg.add("OpenGithubModelsApi")

Quick Start

using OpenGithubModelsApi

# Create a client with your GitHub token
client = GithubModelsClient("YOUR_GITHUB_TOKEN")

# Create a message history
messages = [
    Message(role="system", content="You are a helpful assistant"),
    Message(role="user", content="What is Julia programming language?")
]

# Create an inference request
request = InferenceRequest(
    model="publisher/model-name",  # Replace with an actual model ID
    messages=messages,
    temperature=0.7
)

# Get a response
response = inference_req(client, request)
println(response)

Usage Examples

Basic Chat Completion

using OpenGithubModelsApi

client = GithubModelsClient("YOUR_GITHUB_TOKEN")

# Create a message history
messages = [
    Message(role="system", content="You are a helpful assistant"),
    Message(role="user", content="Explain machine learning in simple terms")
]

# Create an inference request
request = InferenceRequest(
    model="openai/gpt-4.1",  # Example model ID
    messages=messages,
    temperature=0.5,
    max_tokens=200
)

# Get a response (returns just the content by default)
response = inference_req(client, request)
println(response)

# For full response details, use verbose=true
full_response = inference_req(client, request, verbose=true)

Organization-Specific Chat Completion

using OpenGithubModelsApi

client = GithubModelsClient("YOUR_GITHUB_TOKEN")

messages = [
    Message(role="user", content="What models does our organization have access to?")
]

request = InferenceRequest(
    model="publisher/model-name",
    messages=messages
)

# Replace "your-org" with your GitHub organization name
response = org_inference_req(client, request, "your-org")
println(response)

Listing Available Models

using OpenGithubModelsApi

client = GithubModelsClient("YOUR_GITHUB_TOKEN")
models = list_models(client)

# Print available models
for model in models
    println("Model: $(model.name) (ID: $(model.id))")
    println("Publisher: $(model.publisher)")
    println("Summary: $(model.summary)")
    println("---")
end

API Reference

Client Initialization

client = GithubModelsClient(auth_token; 
                           base_url="https://models.github.ai", 
                           api_version="2022-11-28")

Main Functions

• inference_req(client, request; verbose=false) - Creates a standard chat completion
• org_inference_req(client, request, org; verbose=false) - Creates an organization-scoped chat completion
• list_models(client) - Lists all available models

Data Models

The package provides comprehensive type definitions for all API interactions:

• InferenceRequest - Request structure for chat completions
• Message - Represents individual chat messages
• ModelData - Contains metadata about available models
• Various response types for handling API responses

Important Notes

• Streaming is not supported - The API will return an error if stream=true is specified in the request
• All parameters are validated according to GitHub's API specifications
• Temperature values must be between 0-1
• Message roles must be one of: "assistant", "developer", "system", "user"

Documentation

For complete documentation, visit https://imohag9.github.io/OpenGithubModelsApi.jl/stable.

Testing

The package includes a comprehensive test suite that can be run with:

using Pkg
Pkg.test("OpenGithubModelsApi")

License

This project is licensed under the MIT License - see the LICENSE file for details.