AiChat

The Ai Chat page provides a UI, with rich features for interacting with LLM’s on your local machine.

Please remember to enable CORS on your local model runner, ollama or llama.cpp.

This works only on chromium based browsers, enable local App access in site settings, on google Chrome. (Read below…)

The image generation, RAG will be available on the local edition of this tool, shortly. This hosted version of the chat module supports Chat completions and Embedding generation.

Sample screen AI Chat.

Since the requests are Cross Origin Requests, that is, the browser is trying to connect to LLM in your local host, CORS has to be enabled on the LLM provider.

Ollama

For Ollama, use the below commands to allow cors request from binuud.com. On mac and linux, use the below commands.

export OLLAMA_ORIGINS="https://binuud.com"
ollama serve

For windows

setx OLLAMA_ORIGINS "https://binuud.com"
ollama serve

Ollama documentation

Docker Model Runner

Open Docker Desktop.
Navigate to Settings > AI > Click on “Enable Docker Model Runner” Checkbox.
Change the port value to “11434”
Select “CORS Allowed Origins” Dropdown to “https://binuud.com” Docker Settings.

Use latest docker desktop to avoid below issue.
CORS Failing on Docker Model Runner

Browser Support

To utilize this feature, your local machine must host an LLM that’s compatible with the OpenAI API. Please note that Safari may prevent access to http resources from https domains without a valid SSL certificate for your local LLM server.
For using this tool, you might want to try Google Chrome or Firefox instead.

Google Chrome

Go to site settings. Enable Apps on Device Google Settings.

Opensource Availability