https://github.com/RobsonPontin/openai-winrt
An unofficial OpenAI SDK for WinRT applications
https://github.com/RobsonPontin/openai-winrt
chat chatgpt dall-e dalle2 desktop gpt3 openai uwp windows winrt winui winui3
Last synced: about 2 months ago
JSON representation
An unofficial OpenAI SDK for WinRT applications
- Host: GitHub
- URL: https://github.com/RobsonPontin/openai-winrt
- Owner: RobsonPontin
- License: mit
- Created: 2023-01-13T04:28:27.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-09-03T15:33:53.000Z (over 1 year ago)
- Last Synced: 2025-02-06T00:11:27.126Z (3 months ago)
- Topics: chat, chatgpt, dall-e, dalle2, desktop, gpt3, openai, uwp, windows, winrt, winui, winui3
- Language: C++
- Homepage:
- Size: 1.92 MB
- Stars: 7
- Watchers: 6
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# WinRT SDK for OpenAI DALL·E and GPT
A community-maintained (unofficial) OpenAI WinRT SDK that can be used with [UWP (Universal Windows Platform)](https://learn.microsoft.com/en-us/windows/uwp/get-started/universal-application-platform-guide) and/or [WinUI 3 (Desktop)](https://learn.microsoft.com/en-us/windows/apps/winui/winui3/) apps for both C++ and C# languages.
For more information visit https://openai.com/.
## OpenAI Service
It is necessary to initialize the service with an API Key, which can be obtained at https://openai.com/.
The example below show how to get the OpenAI Key by passing your environment variable which contains it.
```cs
var options = new OpenAI.OpenAiOptions();
options.SetOpenAiKeyFromEnvironmentVar("MY_OPEN_AI_API_KEY");
m_openAiService = new OpenAI.OpenAiService(options);
```
## GPT3 APIs
### Chat
Using chat you can go back and forth with the AI models about almost any subject
```cs
var chatRequest = new OpenAI.Chat.ChatRequest();
var chatMessage = new OpenAI.Chat.ChatMessage("user", "What is the most used math formula in the world?");
chatRequest.Messages.Add(chatMessage);
var response = await m_openAiService.RunRequestAsync(chatRequest);
```
### Text Completion
With text completion you can send some text as a prompt, and the model will generate a text completion that attempts to match whatever context or pattern you gave it.
```cpp
winrt::hstring prompt = L"tell me a very nice joke";
auto completionReq = OpenAI::Completion::CompletionRequest{};
completionReq.Prompt(prompt);
completionReq.Model(ModelType::text_davinci_003);
auto response = co_await m_openAiService.RunRequestAsync(completionReq);
```
## DALL·E
### Image Generation
Generate images from a command prompt.
```cpp
winrt::hstring prompt = L"a robot duck walking on the beach";
auto imageReq = OpenAI::Image::ImageCreateRequest{};
imageReq.ImageName(L"ai image");
imageReq.Prompt(prompt);
auto response = co_await m_openAiService.RunRequestAsync(imageReq);
```
### Image Variation
Create image variations from a source image file.
```cpp
Windows::Storage::StorageFile file = GetFileFromPicker();
auto imageReq = OpenAI::Image::ImageVariationRequest{};
imageReq.ImageName(file.Name());
co_await imageReq.SetImageAsync(file);
auto response = co_await m_openAiService.RunRequestAsync(imageReq);
```
### Image Editing
Create a new image from a source image file considering the prompt editing commands.
```cpp
Windows::Storage::StorageFile file = GetFileFromPicker();
WindowS::Storage::StorageFile mask = GetFileFromPicker();
winrt::hstring prompt = L"add a small tiger with shoes";
auto imageReq = OpenAI::Image::ImageEditRequest{};
imageReq.ImageName(file.Name());
imageReq.Prompt(prompt);
co_await imageReq.SetImageAsync(file);
co_await imageReq.SetMaskAsync(mask); // Mask needed to declare which are of the image the editing should be applied
auto response = co_await m_openAiService.RunRequestAsync(imageReq);
```