# OllamaHub

A local Windows proxy that imitates the Ollama API so Visual Studio Copilot Chat's bring-your-own-model feature can be pointed at OpenAI, Anthropic or Ollama-compatible providers instead.

OllamaHub is a local HTTP proxy service that mimics part of the Ollama API on the outside, while forwarding or translating requests to whichever upstream provider a configured model actually uses. Its purpose is to bridge Visual Studio Copilot Chat's bring-your-own-model feature, which expects a local Ollama endpoint, to other model providers.

It implements the small set of Ollama endpoints that bring-your-own-model flow calls, such as listing and describing models, along with OpenAI-compatible and Azure OpenAI-compatible chat completion endpoints. Each configured model can declare which of the OpenAI, Anthropic or Ollama protocols it accepts directly; if the client's protocol is not one of those, the request falls back to a conversion path aimed at Anthropic's API.

Configuration is a single settings file compatible with an existing similar project's provider and model format, run alongside the compiled executable.

Website: https://github.com/mingkuang-Chuyu/OllamaHub