Skip to content

Latest commit

 

History

History

Umbraco.AI.OpenAI

README.md

Umbraco.AI.OpenAI

NuGet

OpenAI provider plugin for Umbraco.AI, enabling integration with OpenAI's GPT models and Azure OpenAI Service.

Features

  • OpenAI API Support - Connect to OpenAI's API (GPT-4, GPT-4o, GPT-3.5-turbo, etc.)
  • Azure OpenAI Support - Connect to Azure OpenAI Service deployments
  • Chat Capabilities - Full support for chat completions with streaming
  • Embedding Capabilities - Generate text embeddings for semantic search
  • Model Configuration - Configure temperature, max tokens, and other model parameters
  • Middleware Support - Compatible with Umbraco.AI's middleware pipeline

Monorepo Context

This package is part of the Umbraco.AI monorepo. For local development, see the monorepo setup instructions in the root README.

Installation

dotnet add package Umbraco.AI.OpenAI

Requirements

  • Umbraco CMS 17.0.0+
  • Umbraco.AI 1.0.0+
  • .NET 10.0

Configuration

After installation, create a connection in the Umbraco backoffice:

  1. Navigate to the AI section
  2. Create a new OpenAI connection
  3. Choose between OpenAI API or Azure OpenAI
  4. Enter your API key or Azure credentials
  5. Create a profile using this connection

OpenAI API

{
    "ApiKey": "sk-...",
    "OrganizationId": "org-..." // Optional
}

Azure OpenAI

{
    "Endpoint": "https://your-resource.openai.azure.com/",
    "ApiKey": "your-azure-api-key",
    "DeploymentName": "your-deployment-name"
}

Supported Models

Chat Models:

  • GPT-4o
  • GPT-4 Turbo
  • GPT-4
  • GPT-3.5 Turbo

Embedding Models:

  • text-embedding-3-large
  • text-embedding-3-small
  • text-embedding-ada-002

Reasoning effort

Chat profiles expose reasoning effort for the o-series, GPT-5 and gpt-6-luna (including dated Luna snapshots). Leave the setting empty to use the model default.

GPT-6 Luna accepts none, low, medium, high, xhigh and max. A stored or selected minimal value maps to low on Luna, following OpenAI's migration guidance, instead of sending an unsupported value or falling back to the medium default.

The dropdown schema is shared across models. xhigh and max are sent unchanged for Luna and the GPT-5.6 family. On other recognized reasoning models, both selections map to high to preserve the intent of choosing more reasoning. Existing models retain their previous handling of none, minimal, low, medium and high. Unknown values and settings on non-reasoning models are skipped. Other GPT-6 variants are outside this declaration because their supported levels differ.

Documentation

License

This project is licensed under the MIT License. See LICENSE.md for details.