OLLAMACHAT is a web application that allows users to chat with various Large Language Models (LLMs). It's built with ASP.NET Core and features a clean, modern architecture. The application supports multiple LLMs and uses a local SQLite database to store chat history.

• Chat Interface: A simple and intuitive web interface for chatting with LLMs.
• Multi-Model Support: Supports multiple LLMs, including models any OpenAI capable provider.
• Chat History: Stores chat history in a local SQLite database.
• Background Job Processing: Uses SignalR to communicate between frontend and backend.
• API Documentation: Includes Swagger for API documentation.
• SSE MCP support: mcp config can be defined in appsettings including passing custom Authorization header.

• .NET 7
• ASP.NET Core
• Entity Framework Core
• SQLite
• SignalR
• Swagger
• Razor Pages
• Minimal APIs (gonna be replaced)
• Mediator
• ModelContextProtocol.Core

The application's configuration is stored in appsettings.json. The following settings can be configured:

• Urls: The URL the application will run on.
• ConnectionStrings:OLLAMACHAT: The connection string for the SQLite database.
• OpenAISettings:ApiKey: Your OpenAI API key.
• OpenAISettings:Models: A list of LLMs to make available in the application.
• McpServers: A list of sse mcp servers to connect to, chat will use them if necessary.

1. Clone the repository:

   git clone https://github.com/Prikalel/OLLAMACHAT.git

2. Navigate to the web project directory:

   cd OLLAMACHAT/Source/OLLAMACHAT.Web

3. Restore dependencies:

   dotnet restore

4. Update the appsettings.json file:
   • Set your OpenAI API key in OpenAISettings:ApiKey.
5. Run the application:

   dotnet run

6. Open your browser and navigate to the URL specified in appsettings.json (default is http://localhost:5001).

PVS-Studio - static analyzer for C, C++, C#, and Java code.