Configure one or more model providers. Each provider entry has a unique name, which is used to select it in other parts of the UI. You can add, modify, and remove providers from the list. The default provider is used when an LLM is required and no other provider is chosen. You can also set the default model for the selected provider.
The current supported model providers are:
The details you will need can be found in the “Keys and Endpoint” section when examining your resource from the Azure portal.
For the API key you can use either KEY1 or KEY2.
You can also find the endpoint via the “Deployments” page in the Azure AI Foundry portal. An example endpoint is: https://docs-test-001.openai.azure.com/.
The API key can be obtained from the Anthropic Console under API Keys .
The endpoint base URL is built-in and does not need to be specified.
The API key can be found in the “API Keys” section of Google AI Studio.
The endpoint base URL is built-in and so does not need to be specified.
You will typically not need to specify an API key. Local models do not need an API key and if you access cloud models then you can configure the API key in Ollama.
If you have Ollama installed on the same machine as ZAP then by default the endpoint will be http://localhost:11434/.
Use this provider for any server that exposes an OpenAI-compatible Chat Completions API.
The Endpoint Base URL field includes suggested values for common servers; you can also enter a custom endpoint.
Each model provider has the following fields:
The Default provider is used by LLM features unless another provider is selected elsewhere. The default provider list includes a None option, which indicates that no provider is configured by default.
The Default model applies to the selected default provider. It is populated from the provider’s model list.
The main ZAP toolbar includes an LLM button that lets you quickly switch the global default provider and model. The dropdown shows one entry for each provider/model combination and indicates which one is currently the default. The None entry clears the default provider and model.
Changing the global default here applies to new chat tabs and any other LLM feature that uses the default. It does not change the provider already selected in an open chat tab — each tab manages its own provider independently.
Each LLM Chat tab has its own provider dropdown in its toolbar. This lets you choose a different provider or model for that conversation without affecting other tabs or the global default. See the LLM Chat help for details.