n8n / Dify / Coze

Connect low-code AI workflow platforms to Chuizi.AI's OpenAI-compatible endpoint for access to 221 models in visual pipelines.

Prerequisites

• A deployed instance of any of the above platforms
• A Chuizi.AI account with an API key (starts with ck-)

n8n Configuration

1. Open the n8n workflow editor
2. Go to Credentials
3. Click Add Credential > select OpenAI API
4. Enter the following:
   • API Key: ck-your-key-here
   • Base URL: https://api.chuizi.ai/v1
5. Save the credential
6. When using an OpenAI node in a workflow, select the credential you just created
7. Enter the model name in the Model field, e.g., anthropic/claude-sonnet-4-6

Dify Configuration

1. Open Dify settings
2. Go to Model Provider
3. Click OpenAI API-compatible
4. Enter the following:
   • API Key: ck-your-key-here
   • API Base URL: https://api.chuizi.ai/v1
5. Add the models you need:
   • Model name: anthropic/claude-sonnet-4-6
   • Model type: LLM
6. After saving, these models will be available in your applications

Coze Configuration

1. Go to the Coze platform plugin settings
2. Find Custom Model Endpoint
3. Configure the API details:
   • Endpoint URL: https://api.chuizi.ai/v1/chat/completions
   • API Key: ck-your-key-here
   • Model: anthropic/claude-sonnet-4-6
4. Save and use in your Bot

Verify the Configuration

Create a simple test in any of the platforms:

1. Add an AI chat node
2. Select a model configured through Chuizi.AI
3. Send a test message
4. Confirm you receive a normal response

Common Issues

Empty Model List in n8n

n8n may not automatically fetch the model list. Manually enter the model name in the provider/model format in the Model field.

Dify Reports Model Unavailable

Make sure you selected the correct model type (LLM) when adding the model. Also verify the API Base URL has no trailing slash.

Workflow Execution Timeout

Some models (like Claude Opus) have longer response times. Increase the timeout in the platform's node settings -- 120 seconds or more is recommended.

Next Steps

• Choose a Model — pick models for workflow tasks
• Function Calling — use tools in your AI workflows
• Structured Output — get JSON for reliable data pipelines