Qwen Portal Setup Tutorial (Beginner-Friendly)

If you want a free, easy first provider to get NextClaw running, Qwen Portal is one of the best starting options.

Note: You do not need a separate pre-existing Qwen account. Open chat.qwen.ai and use any sign-up or sign-in method shown there (for example email, Google, or GitHub). If Qwen changes the login flow later, follow what the page shows.

This tutorial helps you finish four things:

1. Open the Provider settings page in NextClaw
2. Complete Qwen authorization in your browser
3. Enable qwen-portal in NextClaw
4. Send a real test message to confirm it works

Prerequisites

• NextClaw is installed and can be opened.
• Your machine can access chat.qwen.ai.
• You can access chat.qwen.ai and complete sign-up or sign-in with any method shown on the page.

If you use NextClaw from CLI, start the local UI first:

nextclaw start --ui-port 55667

Then open:

• http://127.0.0.1:55667

1) Open the Providers page

Inside NextClaw:

1. Open the Providers page.
2. Find Qwen Portal in the provider list.
3. Open its configuration card.

You should see a built-in authorization section, so you do not need to go hunt for a third-party API key first.

2) Click Authorize in Browser

In the Qwen Portal card:

1. Click Authorize in Browser.
2. NextClaw opens your browser and jumps to the Qwen authorization page.
3. Sign in and complete the authorization steps.

Normally:

• You do not need to manually fill in an API key.
• NextClaw stores the credential automatically after authorization.
• apiBase is also filled with the default value: https://portal.qwen.ai/v1.

3) Return to NextClaw and wait for completion

After the browser flow is done, go back to NextClaw:

1. Keep the current Qwen Portal page open.
2. Wait until the UI shows Authorization completed.
3. Confirm the provider status becomes configured.

If you already signed in with Qwen CLI on this machine, you can also use:

• Import From Qwen CLI

That imports credentials from ~/.qwen/oauth_creds.json, which is handy if you already use the Qwen CLI.

4) Select a model and send a test message

After authorization, do a quick real test:

1. Open the chat page.
2. Select qwen-portal/coder-model in the model picker.
3. Send this message:

Please reply exactly: QWEN-PORTAL-OK

Expected result:

• The model replies with QWEN-PORTAL-OK, or a short equivalent response.

For an even simpler check, you can also send:

Reply with number only: 1+1=

Expected result:

• 2

Who this is for

Qwen Portal is a great fit if you:

• are new to NextClaw and do not want to study API platforms first;
• want to quickly verify chat, tools, or workflows end to end;
• want to start free before switching to another provider later.

Common questions

Why can't I see Qwen Portal?

• Make sure your NextClaw version includes the built-in qwen-portal provider.
• If you are on an older version, upgrade first.

I finished browser login, but NextClaw still does not show success

• Wait a few seconds on the same page; the UI needs a short polling cycle.
• Do not close the Qwen Portal config page immediately.
• If needed, start the browser authorization again.

Do I need to manually fill API Key?

• Usually no.
• NextClaw writes the credential automatically after browser authorization.

Do I need to manually fill API Base?

• Usually no.
• The default is https://portal.qwen.ai/v1.

What does Import From Qwen CLI do?

• If you already signed in with Qwen CLI, NextClaw can reuse that local credential.
• If you never installed Qwen CLI, just ignore this button and use browser authorization.

Related docs

• Tutorial Hub
• Configuration
• Model Selection Guide
• Troubleshooting