Set up Open WebUI with Divyam

Use Open WebUI to test browser-based conversations through the Divyam demo. Open WebUI runs on your machine. Its server sends model requests to the hosted Divyam API.

This guide covers account setup, the connection, conversation tests, and request verification. The installation commands below target macOS and Linux.

Before you start

You need access to your email and an Open WebUI administrator account. For a new local installation, install uv first. uv downloads the Python runtime and runs Open WebUI.

Copy connection values from your console if they differ from these examples.

1 Sign up or sign in to Divyam

1. Open the Divyam demo console.
2. Choose Sign up for a new account, or Sign in for an existing account.
3. Enter your email and complete verification using the code sent to your inbox.
4. Open Credits & Billing and check your available credit.
5. Open API Keys and create a dedicated key, such as open-webui-demo.
6. Copy and privately save the complete key. It is shown once.
7. Open Getting Started and copy the API base URL.
8. Open Models & Pricing and copy an available model ID exactly.

If account access is restricted, contact the demo administrator with your email address. The browser verification code and Divyam API key serve different purposes. Use the API key in Open WebUI.

2 Start Open WebUI

If you already have Open WebUI, open it and continue to step 3. Otherwise, run this command in a terminal:

DATA_DIR="$HOME/.open-webui-divyam" uvx --python 3.11 open-webui@latest serve --host 127.0.0.1 --port 3000

This runs the latest stable package with Python 3.11. DATA_DIR keeps chats and connection settings in a persistent directory on your machine. The port 3000 is the local port chosen for this example.

Wait for startup to finish, then open your local Open WebUI. Keep the terminal running while you test. Press Ctrl+C there when you want to stop Open WebUI.

Complete Open WebUI’s initial account setup. The first account on a fresh installation becomes its administrator. This local account is separate from your Divyam account. For an existing managed installation, ask its administrator to configure the connection.

3 Add the Divyam connection

1. Open Open WebUI’s Settings → Admin → Connections.
2. Find Manage OpenAI API Connections. Enable OpenAI API if it is disabled.
3. Click Add Connection using the plus button.
4. Enter the connection values below.

1. Click Verify Connection and check for a successful connection message.
2. Click Save in the connection dialog. Save the Connections page if your version requires it.
3. Start a new chat and open the model selector.
4. Select the exact Divyam model ID, such as divyam-as/gpt-5.6-sol.

Open WebUI stores this key in its server settings. Protect that installation and its data directory. Users allowed to use this connection can consume its Divyam account’s credit.

Connection verification checks discovery. The conversation test below verifies actual inference and model access. If the model selector is empty, refresh it or reload the page after saving. You can also add an exact model ID under Model IDs, using the plus button before saving.

4 Test conversations through Divyam

Send a short prompt with a recognizable marker:

Reply with exactly: open-webui-switch-check-4827

Check that the answer appears in your chat. Record the test time and selected model. Then test your own representative application inputs:

1. Ask a question with enough detail to assess the answer.
2. Send a follow-up that requires the previous conversation’s context.
3. Confirm that streamed output completes without stalling.
4. Repeat the same inputs with another available Divyam model when comparing results.

Selecting a model in Divyam’s Getting Started page changes its example request. Open WebUI sends the model selected in its own chat. That requested model determines the applicable Divyam plan rates.

Open WebUI can send additional requests for chat titles, tags, and follow-up suggestions. These can use a configured task model. Check those settings if logs show unexpected requests or model IDs.

The connection covers chat completions. Test attachments and tools separately if your application needs them. Reasoning controls have not been verified for this demo route.

5 Verify requests and usage

1. Return to the Divyam demo console. Sign in again if needed.
2. Open Logs in the left navigation.
3. Set Model to the ID selected in Open WebUI.
4. Set Period to include your test time. Leave Status at All statuses.
5. Find the corresponding requests. Times appear in UTC, with newest requests first.
6. Check status 200, token counts, and latency.
7. Open View request and look for your marker or conversation input when content is available.
8. Open View response when available to compare the recorded answer with your chat.
9. Open Usage for request totals and spend, then Credits & Billing for settled charges and remaining credit.

For streamed answers, the console may show Streamed; not stored instead of View response. Compare the request contents and Open WebUI’s displayed answer in that case. Additional background requests mean one visible chat message can produce several log entries.

The logged model identifies the requested Divyam catalog model. It does not identify the backing model that answered. If records are missing, check the signed-in account and filters, then refresh after ingestion.

Troubleshooting and cleanup

When finished, stop the local server with Ctrl+C. For a dedicated demo connection, remove it from Open WebUI and revoke its key in Divyam when no longer needed. To rotate a key, update the saved connection, verify another request, then revoke the old key.

Further reading

• Open WebUI installation
• OpenAI-compatible connections
• Test a custom AI application