Choosing a model

Choose who provides the model during setup. You can change it later in System administration → LLM.

Omnimatch models

Use models provided by Omnimatch without managing your own provider account or model server. Enter a product key with Omnimatch model access. Calls use your Omnimatch credit and require internet access.

Local models

Run a model on your own machine or private network. Enter its OpenAI-compatible endpoint and model name. The model must support tool calling and be reachable from the Docker container.

Ollama is a starting point for running local models. This option supports fully air-gapped operation and does not consume Omnimatch credit.

Your own provider

Connect your OpenAI, Anthropic or compatible provider using your own endpoint, model name and API key. Your provider bills you directly; these calls do not consume Omnimatch credit. Zero data retention depends on your agreement and configuration with that provider.