Skip to main content

Open WebUI

Open WebUI gives you a chat interface in your browser. It runs on your machine, and its server sends model requests to Divyam. The commands below are for macOS and Linux.

Before you start

You need a Divyam account and an API key. If you don't have them yet, follow Getting Started first.

Replace the shortened divyam-sw-v1-... shown in examples with your complete API key.

The examples use divyam-as/gpt-6-sol. To choose another model, use this guide’s model selector when one is shown, or copy its current ID from Models and pricing. Check its context limit and supported features.

Set it up with an AI agent

Paste this into Claude Code, Codex or another coding agent. It installs and configures everything, checks it works, and tells you how to start.

Prompt
Set up Open WebUI with Divyam using https://docs.divyam.ai/divyam-setup.md

Your key never goes in the chat. The agent asks for it in a private prompt and saves it here, so later setups reuse it:

SystemWhere the key is saved
macOSLogin Keychain
Linux with a desktop keyringThe desktop keyring
Linux without a keyringA file only your user can read
WindowsA file encrypted for your Windows user account

Setting up an application also saves the key in that application's own settings.

You also need an Open WebUI admin account. For a new local install, install uv first. uv downloads Python and runs Open WebUI for you.

1. Start Open WebUI​

If you already run Open WebUI, open it and go to step 2. Otherwise, run:

DATA_DIR="$HOME/.open-webui-divyam" uvx --python 3.11 open-webui@0.11.4 serve --host 127.0.0.1 --port 3000

This runs Open WebUI 0.11.4, the version Divyam tested these steps with, on Python 3.11. DATA_DIR keeps your chats and settings in a folder on your machine. Leave the terminal running, and press Ctrl+C in it when you want to stop Open WebUI.

When startup finishes, open http://127.0.0.1:3000 and create the first account. On a fresh install, the first account becomes the Open WebUI admin. This account is separate from your Divyam account.

2. Add the Divyam connection​

  1. Open Settings → Admin → Connections.

  2. Under Manage OpenAI API Connections, turn on OpenAI API if it is off.

  3. Click the plus button to Add Connection, and enter these values:

    FieldWhat to enter
    URLhttps://api.preview.divyam.ai/v1
    AuthBearer
    API KeyYour Divyam API key on its own, without the word Bearer in front
    API TypeChat Completions
    Provider (under Advanced)Default
    Prefix IDLeave empty
    Model IDsLeave empty so Open WebUI lists models from Divyam
  4. Click the refresh icon next to the URL to verify the connection, and wait for the success message.

  5. Click Save in the dialog. Save the Connections page too if your version asks.

Open WebUI Edit Connection dialog with the Divyam URL, Bearer auth with the key masked, and API type Chat Completions

warning

Open WebUI stores the key in its server settings. Protect the install and its data folder: anyone who can use this connection spends your Divyam credit.

3. Chat with a Divyam model​

Start a new chat, open the model picker, and select your Divyam model, such as divyam-as/gpt-6-sol.

Open WebUI model picker listing the Divyam models

If the list is empty, reload the page. You can also add the exact model ID under Model IDs in the connection, using the plus button, then save.

Send a short message:

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

The answer appears in the chat. Then send a follow-up that depends on the first message to confirm the conversation carries over.

Open WebUI chat where divyam-as/gpt-6-sol replies with the check phrase

note

Open WebUI may send extra requests for chat titles, tags and follow-up suggestions, using its task model setting. One chat message can produce several requests in Divyam.

Check your requests​

Open the Divyam console and select Logs. Your requests appear there within a few seconds, newest first, with their status, token counts and latency. Usage and Credits & Billing catch up within a few minutes.

Local cost estimates

Some clients display zero when model prices are missing. Check Credits & Billing for Divyam charges. Local dollar limits depend on the client's price and usage data.

For filters, cached-token counts and the requested-versus-routed view, see Logs and usage.

Troubleshooting​

SymptomWhat to do
The local page doesn't openCheck the terminal and wait for startup to finish.
Port 3000 is in usePick another free port with --port and open that port in your browser.
No Admin settingsSign in with the Open WebUI admin account, or ask whoever runs the install.
No models in the pickerCheck the URL ends in /v1, verify the connection, save, and reload.
HTTP 401Check you pasted the whole key with Bearer auth, and that the key is still active.

For other errors, see Troubleshooting.

Clean up​

Stop the local server with Ctrl+C. To disconnect, remove the Divyam connection in Settings → Admin → Connections and revoke the key in API keys if nothing else uses it. To rotate the key, update the saved connection, send one message, then revoke the old key.