Open WebUI sends each chat turn to the model shown in the chat model picker. When Ollama provides several local models, that choice determines which model handles the next message.
The picker lists models available to your account, which can include local and cloud providers. An Ollama model must already be downloaded and reachable through a saved connection before it can be selected.
A new chat keeps the test separate from an existing conversation. The model name beside the reply identifies the entry used for that response; a model's own claim about its identity is not a reliable check.
Steps to select an Ollama model in Open WebUI:
- Open a new chat with New Chat in Open WebUI.

- Open the model picker at the lower right of the message box.
The Search a model field filters long lists of available models.
- Select the intended Ollama model from the list.
company-chat:latest is a sample model name; your server may use a different name or tag. If the model is missing, the saved connection, refreshed model list, and access granted to your account determine whether it appears.
Related: How to connect Ollama to Open WebUI
Related: How to refresh the Open WebUI model list - Enter a short test prompt in the message box.
What is 2 + 2? Answer in one short sentence.
- Click Send message to submit the prompt.

- Confirm that a reply appears under the selected Ollama model name.
The response label should match the picker selection, such as company-chat:latest. A model label without an answer does not show that generation succeeded.
Related: How to fix Open WebUI localhost connection errors in Docker