Integration Guide

n8n + AIPower

Use one OpenAI-compatible endpoint in n8n to run workflow automation across GPT, Claude, Gemini, DeepSeek, Qwen, GLM, Kimi, and Doubao.

Get API key Try demo

Support automation

Turn tickets, chats, and form submissions into routed replies with model-level cost control.

Sales ops

Summarize leads, enrich CRM notes, classify intent, and draft follow-ups from one API key.

Internal AI agents

Give operations teams GPT, Claude, Gemini, DeepSeek, Qwen, GLM, Kimi, and Doubao behind one billing layer.

Setup option A: OpenAI-compatible config

Create an AIPower API key from your dashboard.
In n8n, add or edit your OpenAI credential or OpenAI Chat Model node.
Use AIPower as the OpenAI-compatible endpoint when your n8n version exposes a Base URL field:
Base URL: https://api.aipower.me/v1
API Key: your-aipower-key
Model: deepseek/deepseek-chat
For n8n's OpenAI node, choose the chat-completions style operation when available. AIPower is designed around OpenAI-compatible chat completions.

Setup option B: HTTP Request fallback

If your n8n build does not expose a custom Base URL field for the OpenAI model node, use the HTTP Request node. This works anywhere n8n can make an HTTPS POST request.

Method: POST

URL: https://api.aipower.me/v1/chat/completions

Headers: Authorization: Bearer your-aipower-key

{
  "model": "deepseek/deepseek-chat",
  "messages": [
    {
      "role": "system",
      "content": "You classify support tickets for an operations team."
    },
    {
      "role": "user",
      "content": "={{ $json.message }}"
    }
  ],
  "temperature": 0.2
}

Recommended models for n8n workflows

Workflow job	AIPower model	Why
Customer support triage	deepseek/deepseek-chat	Classify tickets, draft replies, and route complex cases to a stronger model.
Chinese pre-sales bot	qwen/qwen-plus	Answer product, shipping, refund, and policy questions in Chinese workflows.
Low-cost intent routing	zhipu/glm-4-flash	Use for tags, short JSON, lead scoring, and simple workflow branching.
Complex troubleshooting	anthropic/claude-sonnet	Use after top-up for long context, support escalations, and technical reasoning.
Long document summary	google/gemini-2.5-flash	Good for policies, docs, conversation logs, and long workflow inputs.

Why AIPower fits n8n teams

One endpoint for many workflows. Keep n8n automation simple while switching between China and global model families.
Payment flexibility. Use card, WeChat Pay, or Alipay instead of opening separate provider billing accounts.
Better automation economics. Use low-cost models for classification and summaries, then reserve GPT, Claude, or Gemini for harder branches.
Team/BYOK upgrade path. Add shared budgets, invoices, logs, and provider-key routing when workflows move from prototype to production.

Connect your next n8n AI workflow

10 trial calls. One OpenAI-compatible endpoint. China + global models.

Get API key