A starting point for interacting with small LLMs and embeddings as part of basic chat services.
Tested with Microsoft Foundry OpenAI SDK endpoints
In folder: src/fastapi, copy .env.example as .env, add api key and endpoint.
Tools: uv, python3.14
cd src/fastapi
uv sync
uv run uvicorn main:appIn folder src/webapi, copy appsettings.json as appsettings.Development.json, add api key and endpoint.
Tools: dotnet + .NET 10 SDK
cd src/webapi
dotnet restore
dotnet runTools: node, npm
cd src/javascript
npm install
npm run devOpen browser at http://localhost:8080 and experiment!
Use the client UI to add one or both backends:
| Backend | URL |
|---|---|
| WebAPI | http://localhost:8090/ask |
| FastAPI | http://localhost:8000/ask |