Unreal Engine Generative AI Support Plugin

Warning

This plugin is currently under development and not ready for production use.

This project aims to build a long-term support (LTS) plugin for various cutting-edge LLM/GenAI models & foster a community around it. It currently includes OpenAI's GPT4o & GPT4o-mini for Unreal Engine 5.1 or higher, with plans to add Claude Sonnet 3.5, real-time APIs, Deepseek, Gemini, & Grok 2 APIs soon. Will only focus on the APIs that can be useful for game development and interactive experiences. Any suggestions and contributions are welcome. Currently working on OpenAI API support with real-time chat/audio completions.

Current Progress:

• OpenAI API Support:
  • OpenAI Chat API ✅
    • gpt-4o Model ✅
    • gpt-4o-mini Model ✅
    • o1-mini Model 🚧
    • o1 Model 🚧
  • OpenAI DALL-E API 🛠️
  • OpenAI Vision API 🚧
  • OpenAI Realtime API 🛠️
  • OpenAI Structured Outputs ✅
  • OpenAI Text-To-Speech API 🚧
  • OpenAI Whisper API 🚧
• Anthropic Claude API Support:
  • Claude Chat API 🚧
    • claude-3-5-sonnet-latest Model 🚧
    • claude-3-5-haiku-latest Model 🚧
    • claude-3-opus-latest Model 🚧
  • Claude Vision API 🚧
• XAI (Grok) API Support:
  • XAI Chat Completions API 🚧
    • grok-beta Model 🚧
    • grok-beta Streaming API 🚧
  • XAI Image API 🚧
• Google Gemini API Support:
  • Gemini Chat API 🚧
    • gemini-2.0-flash-exp Model 🚧
    • gemini-1.5-flash Model 🚧
    • gemini-1.5-flash-8b Model 🚧
  • Gemini Multi-Modal API 🚧
• Meta AI API Support:
  • Llama Chat API 🚧
    • llama3.3-70b Model 🚧
    • llama3.1-8b Model 🚧
  • Multi-Modal Vision API 🚧
    • llama3.2-90b-vision Model 🚧
  • Local Llama API 🚧
• Deepseek API Support:
  • Deepseek Chat API 🚧
    • deepseek-chat (DeepSeek-V3) Model 🚧
• API Key Management
  • Cross-Platform Secure Key Storage ✅
  • Encrypted Key Storage 🛠️
  • Cross Platform Testing 🚧
  • Build System Integration 🛠️
  • Keys in Build Configuration 🛠️
• Unreal Engine Integration
  • Blueprint Support 🛠️
  • C++ Support 🛠️
  • C++ Latent Functions For Blueprints 🛠️
  • Packaged Build Support 🛠️
• Plugin Documentation 🛠️
• Plugin Examples 🚧
• Version Control Support
  • Perforce Support 🚧
  • Git Submodule Support ✅
• LTS Build Support 🚧
• Lightweight Plugin (In Builds)
  • No External Dependencies ✅
    • Build Flags to enable/disable APIs 🚧
    • Submodules per API Organization 🚧
• Testing
  • Automated Testing 🚧
  • Build Testing 🚧
  • Different Platforms 🚧
  • Different Engine Versions 🚧

Quick Links:

• OpenAI API Documentation
• Anthropic API Documentation
• XAI API Documentation
• Google Gemini API Documentation
• Meta AI API Documentation
• Deepseek API Documentation

Setting API Key:

For Editor:

Set the environment variable PS_OPENAIAPIKEY to your API key. In windows you can use:

setx PS_OPENAIAPIKEY "your api key"

In Linux/MacOS you can use:

export PS_OPENAIAPIKEY="your api key"

For Packaged Builds:

Still in development..

Adding the plugin to your project:

With Git:

1. Add the Plugin Repository as a Submodule in your project's repository.

   git submodule add https://github.com/prajwalshettydev/UnrealGenAISupport Plugins/GenerativeAISupport

2. Regenerate Project Files: Right-click your .uproject file and select Generate Visual Studio project files.

3. Enable the Plugin in Unreal Editor: Open your project in Unreal Editor. Go to Edit > Plugins. Search for the Plugin in the list and enable it.

4. For Unreal C++ Projects, include the Plugin's module in your project's Build.cs file:

   PrivateDependencyModuleNames.AddRange(new string[] { "GenerativeAISupport" });

With Perforce:

Still in development..

Fetching the Latest Plugin Changes:

With Git:

you can pull the latest changes with:

cd Plugins/GenerativeAISupport
git pull origin main

Or update all submodules in the project:

git submodule update --recursive --remote

With Perforce:

Still in development..

Usage:

OpenAI:

Function GetGenerativeAIApiKey by default responds with OpenAI API key, that you have securely set in the local environment variable

1. Chat:

   C++ Example:

   void SomeDebugSubsystem::CallGPT(const FString& Prompt, 
       const TFunction<void(const FString&, const FString&, bool)>& Callback)
   {
       FGenChatSettings ChatSettings;
       ChatSettings.Model = TEXT("gpt-4o-mini");
       ChatSettings.MaxTokens = 500;
       ChatSettings.Messages.Add(FGenChatMessage{ TEXT("system"), Prompt });
   
       FOnChatCompletionResponse OnComplete = FOnChatCompletionResponse::CreateLambda(
           [Callback](const FString& Response, const FString& ErrorMessage, bool bSuccess)
       {
           Callback(Response, ErrorMessage, bSuccess);
       });
   
       UGenOAIChat::SendChatRequest(ChatSettings, OnComplete);
   }

   Blueprint Example:

   

2. Structured Outputs: C++ Example 1: Sending a custom schema json directly to function call

   FString MySchemaJson = R"({
   "type": "object",
   "properties": {
       "count": {
           "type": "integer",
           "description": "The total number of users."
       },
       "users": {
           "type": "array",
           "items": {
               "type": "object",
               "properties": {
                   "name": { "type": "string", "description": "The user's name." },
                   "heading_to": { "type": "string", "description": "The user's destination." }
               },
               "required": ["name", "role", "age", "heading_to"]
           }
       }
   },
   "required": ["count", "users"]
   })";
   
   UGenAISchemaService::RequestStructuredOutput(
       TEXT("Generate a list of users and their details"),
       MySchemaJson,
       [](const FString& Response, const FString& Error, bool Success) {
          if (Success)
          {
              UE_LOG(LogTemp, Log, TEXT("Structured Output: %s"), *Response);
          }
          else
          {
              UE_LOG(LogTemp, Error, TEXT("Error: %s"), *Error);
          }
       }
   );

   C++ Example 2: Sending a custom schema json from a file

   #include "Misc/FileHelper.h"
   #include "Misc/Paths.h"
   FString SchemaFilePath = FPaths::Combine(
       FPaths::ProjectDir(),
       TEXT("Source/:ProjectName/Public/AIPrompts/SomeSchema.json")
   );
   
   FString MySchemaJson;
   if (FFileHelper::LoadFileToString(MySchemaJson, *SchemaFilePath))
   {
       UGenAISchemaService::RequestStructuredOutput(
           TEXT("Generate a list of users and their details"),
           MySchemaJson,
           [](const FString& Response, const FString& Error, bool Success) {
              if (Success)
              {
                  UE_LOG(LogTemp, Log, TEXT("Structured Output: %s"), *Response);
              }
              else
              {
                  UE_LOG(LogTemp, Error, TEXT("Error: %s"), *Error);
              }
           }
       );
   }

Contribution Guidelines:

Project Structure:

References:

• Env Var set logic from: OpenAI-Api-Unreal by KellanM