Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.platform.qubrid.com/llms.txt

Use this file to discover all available pages before exploring further.

OpenClaw is a personal AI assistant you run on your own devices. Connect it to Qubrid AI to use Qubrid models through OpenClaw’s onboarding flow and interact via the terminal, web UI, or messaging channels.
Prerequisites:

​
Install OpenClaw

OpenClaw supports multiple installation methods. See the OpenClaw GitHub repository for Docker, Nix, and other setup options. This guide uses npm (you can also use pnpm add -g openclaw@latest).
1

Install OpenClaw globally

Open a command prompt or terminal and run:
npm install -g openclaw@latest
Or with pnpm:
pnpm add -g openclaw@latest
2

Start onboarding

Run the onboarding command with the daemon installer:
openclaw onboard --install-daemon
This launches the OpenClaw onboarding wizard and installs the gateway service so OpenClaw stays running in the background.

​
Configure Qubrid AI in OpenClaw

1

Review the security notice

Read through the security notice carefully. Choose Yes or No based on your preference.Use the cursor keys to move between options. Selecting No stops onboarding-choose Yes to continue.
2

Select Quickstart onboarding mode

When prompted for Onboarding Mode, choose Quickstart (you can use Manual later for advanced configuration).
3

Select Custom provider

In the Model Provider section, select Custom provider.Custom providers support any OpenAI- or Anthropic-compatible API. Qubrid endpoints are OpenAI-compatible, so this is the right choice for Qubrid AI.
4

Enter the Qubrid API base URL

When prompted for the API base URL, clear the default value and enter:
https://platform.qubrid.com/v1
This is the Qubrid OpenAI-compatible API base URL.
5

Paste your Qubrid API key

When asked how you want to provide the API key, select Paste API Key now.
  1. Open the Qubrid AI Platform and create an API key. Follow the steps in How to generate an API Key.
  2. Copy the key and paste it into the onboarding prompt.
Your API key is shown only once for security. Store it somewhere safe and secure before closing the dialog. If you lose it, you must generate a new key.
6

Select OpenAI-compatible endpoint

When prompted for endpoint compatibility, select OpenAI Compatible.
7

Enter the model ID

Enter the model ID for the Qubrid model you want to use.For example, to use GPT OSS 120B, enter:
openai/gpt-oss-120b
On the Qubrid platform, open the model page-the model ID is shown beside the model name. Click the copy button next to the ID to copy it exactly.
8

Wait for verification

OpenClaw verifies the connection to your Qubrid endpoint. This may take a moment.
9

Set endpoint ID and model alias (optional)

After verification:
  • Endpoint ID - A custom label you choose (for example qubrid or qubrid-gpt-oss-120b).
  • Model alias - Optional display name for the model in OpenClaw. Set it however you prefer, or leave it blank.
10

Configure or skip channels

OpenClaw asks you to select a channel, such as Telegram, WhatsApp, Discord, Slack, Google Chat, or Microsoft Teams.
  • Configure now - Follow the extra steps to link your account on the channel you choose.
  • Skip for now - You can add channels later and use OpenClaw from the TUI or web UI in the meantime.
11

Configure or skip search providers

You can connect search providers such as Brave Search, Gemini, Grok (xAI), Kimi, or Perplexity.For this guide, skip search provider setup for now. You can configure them later if needed.
12

Configure or skip skills

When asked whether to configure skills, select No for now, or configure them if you want OpenClaw to use bundled or custom skills immediately.
13

Configure hooks

OpenClaw asks whether to enable hooks. Enable all hooks, select only the ones you need, or skip based on your preference.
14

Confirm gateway installation

After onboarding completes, the gateway service is installed and running.
15

Launch OpenClaw

Back in the command prompt, choose how you want to hatch your bot:
  • TUI (recommended) - Terminal user interface
  • Web UI - Browser-based interface
OpenClaw opens on localhost. Your default Qubrid model appears as the selected model-you are ready to explore.
Your Qubrid integration with OpenClaw is now configured. You can chat with your assistant in the TUI or web UI, and add messaging channels whenever you are ready.

​
Things you can do with OpenClaw

Once setup is complete, try these examples:
  • Chat in the TUI or web UI - Ask questions, draft content, or debug code using your Qubrid model.
  • Run agent tasks - Use openclaw agent to run multi-step tasks with configurable thinking depth.
  • Connect messaging apps - Talk to your assistant on WhatsApp, Telegram, Discord, Slack, and other supported channels.
  • Send messages from the CLI - Use openclaw message send to send messages through a connected channel.
  • Add skills and automation - Enable skills, cron jobs, and hooks for recurring or event-driven workflows.
  • Switch models - Change the default model or add more Qubrid models through OpenClaw’s model configuration.
For more commands and features, see the OpenClaw documentation.

​
Monitor usage on Qubrid

Keep an eye on your usage and costs on the Qubrid platform.
  • Review Inference Logs to see request history and debug issues.
  • Check your credit balance before long-running or high-volume sessions.
  • Rotate API keys if you suspect a key was exposed.

​
Configuration reference

Use these values when connecting OpenClaw to Qubrid during onboarding:
SettingValue
ProviderCustom provider
API base URLhttps://platform.qubrid.com/v1
Endpoint compatibilityOpenAI Compatible
Example model IDopenai/gpt-oss-120b
API keyYour Qubrid API key
Model IDs vary by model-always copy the ID from the model page on the Qubrid platform.

​
Frequently Asked Questions (FAQ)

You must accept the security notice to continue. Use the cursor keys to move between options and select Yes.Selecting No stops onboarding. Choose Yes to proceed with setup.
Confirm the following:
  • API base URL is exactly https://platform.qubrid.com/v1
  • API key is valid and pasted correctly (generate a new key if needed from Manage API Keys)
  • Model ID matches the value on the Qubrid model page (for example openai/gpt-oss-120b)
  • Endpoint compatibility is set to OpenAI Compatible
Ensure model verification completed successfully during onboarding. Check that your endpoint ID and model alias were saved.If the model still does not appear, restart or check the gateway:
openclaw gateway status
Check your network connection and confirm the Qubrid base URL is exactly https://platform.qubrid.com/v1. Review the model in the AI Playground for expected latency.Use Inference Logs to see whether requests are failing or taking longer than expected on the platform side.
Open Inference Logs on the Qubrid platform to review API usage, monitor spend, and troubleshoot failed requests.

Go to OpenClaw.ai

If you have any questions related to OpenClaw, this is your place to be