Skip to content

README.md

Latest commit

 

History

History
116 lines (77 loc) · 3.51 KB

README.md

File metadata and controls

116 lines (77 loc) · 3.51 KB

Aspire.Hosting.GitHub.Models library

Provides extension methods and resource definitions for an Aspire AppHost to configure GitHub Models.

Getting started

Prerequisites

  • GitHub account with access to GitHub Models
  • GitHub personal access token with appropriate permissions (models: read)

Install the package

In your AppHost project, install the Aspire GitHub Models Hosting library with NuGet:

dotnet add package Aspire.Hosting.GitHub.Models

Usage example

Then, in the AppHost.cs file of AppHost, add a GitHub Model resource and consume the connection using the following methods:

var builder = DistributedApplication.CreateBuilder(args);

var chat = builder.AddGitHubModel("chat", "openai/gpt-4o-mini");

var myService = builder.AddProject<Projects.MyService>()
                       .WithReference(chat);

The WithReference method passes that connection information into a connection string named chat in the MyService project.

In the Program.cs file of MyService, the connection can be consumed using a client library like Aspire.Azure.AI.Inference:

Inference client usage

builder.AddAzureChatCompletionsClient("chat")
       .AddChatClient();

Configuration

The GitHub Model resource can be configured with the following options:

API Key

The API key can be set as a configuration value using the default name {resource_name}-gh-apikey or the GITHUB_TOKEN environment variable.

Then in user secrets:

{
    "Parameters": 
    {
        "chat-gh-apikey": "YOUR_GITHUB_TOKEN_HERE"
    }
}

Furthermore, the API key can be configured using a custom parameter:

var apiKey = builder.AddParameter("my-api-key", secret: true);
var chat = builder.AddGitHubModel("chat", "openai/gpt-4o-mini")
                  .WithApiKey(apiKey);

Then in user secrets:

{
    "Parameters": 
    {
        "my-api-key": "YOUR_GITHUB_TOKEN_HERE"
    }
}

Connection Properties

When you reference a GitHub Model resource using WithReference, the following connection properties are made available to the consuming project:

GitHub Model

The GitHub Model resource exposes the following connection properties:

Property Name Description
Uri The GitHub Models inference endpoint URI, with the format https://models.github.ai/inference
Key The API key (PAT or GitHub App token) for authentication
Model The model identifier for inference requests, for instance openai/gpt-4o-mini
Organization The organization attributed to the request (available when configured)

Aspire exposes each property as an environment variable named [RESOURCE]_[PROPERTY]. For instance, the Uri property of a resource called db1 becomes DB1_URI.

Available Models

GitHub Models supports various AI models. Some popular options include:

  • openai/gpt-4o-mini
  • openai/gpt-4o
  • deepseek/DeepSeek-V3-0324
  • microsoft/Phi-4-mini-instruct

Check the GitHub Models documentation for the most up-to-date list of available models.

Additional documentation

Feedback & contributing

https://github.com/dotnet/aspire