Skip to main content
OpenClaw is a personal AI assistant that you run on your own devices. It answers you on the channels you use every day — WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Microsoft Teams, and more. Your conversations, memory, and data stay on your own machine, not on a third-party server. If you want a personal, single-user assistant that feels local, fast, and always-on, this is it. OpenClaw now supports a wide range of AI model providers — all you need is your own API key to get started. In this guide, we’ll walk you through how to connect SiliconFlow’s APIs to OpenClaw, so you can power your personal assistant with 100+ SOTA AI models in just a few steps.

​
Step 1: Prerequisites

​
1.1 Get Your SiliconFlow API Key

Before you begin, ensure you have a valid SiliconFlow account: You will need this key when configuring OpenClaw.

​
1.2 Node 22 or newer

To verify your Node version: 
node --version
If your version is lower than 22, please upgrade Node.js before proceeding.

​
Step 2: Install OpenCla

OpenClaw Recommended Installation Method

​
2.1 macOS/Linux:

curl -fsSL https://openclaw.ai/install.sh | bash

​
2.2 Windows (PowerShell):

iwr -useb https://openclaw.ai/install.ps1 | iex
For alternative installation methods and system requirements, refer to the official OpenClaw documentation.

​
Step 3: Configure SiliconFlow APIs

​
3.1 Initialize Configuration

After installation, run the following CLI in your terminal: 
openclaw setup
This creates the openclaw.json configuration file.

​
3.2 Locate the Configuration File

On Windows: 
%USERPROFILE%\.openclaw\openclaw.json
On macOS/Linux: 
~/.openclaw/openclaw.json
Open the file in your preferred editor.

​
3.3 Add SiliconFlow as a Model Provider

Merge the following configuration into the "models" section of your openclaw.json file. 
"models": {
    "providers": {
      "siliconflow": {
        "baseUrl": "https://api.siliconflow.com/v1",
        "apiKey": "YOUR_SILICONFLOW_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "deepseek-ai/DeepSeek-V3",
            "name": "DeepSeek-V3",
            "reasoning": false,
            "input": [
              "text"
            ],
            "cost": {
              "input": 0,
              "output": 0,
              "cacheRead": 0,
              "cacheWrite": 0
            },
            "contextWindow": 200000,
            "maxTokens": 8192
          }
        ]
      },
If you already have custom model configurations, merge only the relevant provider block without overwriting existing entries.

​
3.4 Register the Model in the Agents Section

Add the following entry under the "agents" section: 
"siliconflow/deepseek-ai/DeepSeek-V3"
Important:
  • The model identifier must match the ID defined in the provider configuration.
  • It is recommended to first add the model under the “fallbacks” section to prevent runtime errors caused by misconfiguration.
Image

​
3.5 Verify Model Registration

Run: 
openclaw models list
If the model appears in the list, the configuration was successful.
Modellist
By updating the "models"and "agents" sections, you can add more models supported by SiliconFlow With SiliconFlow configured, you can now connect OpenClaw to a messaging platform. In the following section, we will use Discord as an example to demonstrate how to deploy an OpenClaw bot on Discord using SiliconFlow APIs. For detailed configuration guides on additional channels — including Telegram, WhatsApp, Slack, Signal, iMessage, Microsoft Teams, and more — please visit the official documentation at https://docs.openclaw.ai/channels.

​
Example: Run OpenClaw in Discord

​
Create a Discord Application

  • Go to the Discord Developer Portal.
  • Click New Application.
  • Enter a bot name (e.g., OpenClaw (SiliconFlow API)) and then Create.
Newapp

​
Configure the Bot

Navigate to the Bot section.

​
Reset and Copy Bot Token

  • Click Reset Token.
  • Copy the token and store it securely. You will need it during the OpenClaw onboarding process.
Image

​
Enable Required Intents

Enable the following features:
  • Presence Intent
  • Server Members Intent
  • Message Content Intent
Click Save Changes.
Image

​
Configure OAuth2

Go to OAuth2 → URL Generator.

​
Scopes

Enable the bot and applications.commands scopes for basic functionality.
Image

​
Bot Permissions

  • Select the necessary permissions required for your use case.
  • Copy the generated URL at the bottom of the page.
  • Open the URL in your browser and:
    • Select the target Discord server
    • Confirm authorization
    • Add the bot to your server
Image

​
Onboard OpenClaw

  • In your terminal, run:
openclaw onboard
  • During the openclaw onboard process:
    • Confirm that you understand OpenClaw operates as a personal-by-default assistant, then select Yes to proceed.
    • Choose QuickStart as the onboarding mode.
    • Select Use existing values to retain the previously configured workspace, model, and gateway settings. This ensures that your previously configured SiliconFlow model remains active without requiring reconfiguration.
    Image
  • Select Skip for now in the Model/auth provider step.
  • Select siliconflowin the Filter models by provider step.
This instructs OpenClaw to use the SiliconFlow provider block defined in openclaw.json. If SiliconFlow has already been configured correctly, it will appear in the provider list automatically.
Modelauthprovider
  • Under Default model, choose Keep current (siliconflow/deepseek-ai/DeepSeek-V3) which confirms that OpenClaw will use the previously configured SiliconFlow model.
  • Under Select channel (QuickStart), choose Discord (Bot API).
This step links the selected SiliconFlow model to the Discord channel, enabling AI-powered conversations within your Discord server.
Defaultmodel
  • Paste your previously prepared Discord bot token into the input field.
  • In the Configure Skills step, select Yes to manually configure skills, or select No to skip — the system will automatically enable a default set of skills on startup.
  • Choose how the bot should interact with your Discord channels.
Choosechannels
  • Pairing Your Discord Bot with OpenClaw
    • Open the Discord server that your bot has been invited to.
    • Locate the bot you created in the member list and send it a direct message.
    • The bot will automatically respond with a pairing code.
  • Approve the Pairing via CLI
On the server where OpenClaw is running, execute the following command in your terminal, replacing XXXXXX with the pairing code returned by your bot: 
openclaw pairing approve discord XXXXXX
Once the command is executed successfully, the Discord bot will be authenticated and linked to your OpenClaw instance.
Finalresults
Whether you’re building a productivity assistant for daily task management, a knowledge bot for your team’s Discord server, or a personal AI companion available across messaging platforms, OpenClaw powered by SiliconFlow APIs gives you the flexibility to deploy cutting-edge models on your own terms.