This guide helps you set up the language models available on the Lehmus AI platform in Open WebUI. The connection is created by starting Open WebUI with Docker and providing it with the OpenAI-compatible API endpoint for Lehmus AI, as well as your own API key created on the Lehmus AI platform.

The Lehmus AI platform also includes its own chat interface. Open WebUI provides an alternative chat interface, which you can use, for example, when you want to access Lehmus AI language models in a separate Open WebUI environment.

 

Before you start, make sure that…

• You have access to the Lehmus AI platform. Instructions for getting access are available on the guide page Lehmus AI: A Secure Generative AI Platform.
• You have a valid Lehmus AI API key.
• Docker is installed on your computer.
• You have access to a terminal or command line.
• The selected port, 3000, is available. You can change the port number if needed.

 

You will also need the following information

Open WebUI connects to the Lehmus AI platform through an OpenAI-compatible API. Before running the command, you will need the following information:

• Lehmus AI API address: https://api.lehmus-ai.oulu.fi/v1/
• Your API key, which you have created on the Lehmus AI platform. Replace the example text in the next section with your own key.
• Local address where Open WebUI opens in the browser: http://localhost:3000

 

Start Open WebUI with Docker

Once your access, API key, and Docker are ready, start Open WebUI with the following command:

docker run -d \
–name open-webui \
-p 3000:8080 \
-v open-webui:/app/backend/data \
-e OPENAI_API_BASE_URL=”https://api.lehmus-ai.oulu.fi/v1/” \
-e OPENAI_API_KEY=”cm_api_key” \
–restart always \
ghcr.io/open-webui/open-webui:main

Replace cm_api_key with your own Lehmus AI platform API key before running the command. The command downloads and starts Open WebUI automatically. It also stores Open WebUI data so that settings and user information are kept even after restarting your computer. If port 3000 is already in use, change the port number in the command above to another available port.

 

How to open Open WebUI

1. Open your browser and go to http://localhost:3000.
2. Create an Open WebUI user if needed, or sign in.
3. Open the model menu in the top-left corner and check that the Lehmus AI language models are available.
4. Select a model and send a short test message, for example: Hi, is the connection working?

 

Notes

• Do not share your API key with other users.
• If port 3000 is already in use, change the port mapping in the command, for example to -p 3001:8080, and open the corresponding address in your browser.
• If the models do not appear in Open WebUI, check the API key and API address.
• If Open WebUI has already been used on the same computer, previous settings may affect which models you see. If needed, check in the Open WebUI admin view that the Lehmus AI API address and settings are correct.

 

What if the model names look unclear?

Some Lehmus AI language models may appear in Open WebUI with unclear or difficult-to-read names because they are shown based on their model id values. You can change the displayed model names in Open WebUI to make them clearer.

1. On the Open WebUI home page, click the model name.
2. Next to the model, select the three dots, More.
3. Select Edit.
4. When the Model settings view opens, click the title field at the top and change the displayed model name to the correct one.
5. Compare the model id with the models and their ID information on the Models tab of the Lehmus AI platform. This helps you find which clear model name belongs to each unclear model id.

« Back