Switch from OpenAI API to AIPower in 5 Minutes: Migration Guide

If you're currently using the OpenAI API, switching to AIPower takes exactly two lines of code. You keep your existing OpenAI SDK, your existing code, and your existing workflow. You gain access to 16 models (including Chinese AI) and potentially save 90% on costs.

The 2-Line Migration

Before (OpenAI direct):

from openai import OpenAI

client = OpenAI(
    api_key="sk-your-openai-key",
    # base_url defaults to api.openai.com
)

After (AIPower):

from openai import OpenAI

client = OpenAI(
    base_url="https://api.aipower.me/v1",  # ADD this line
    api_key="your-aipower-key",             # CHANGE this line
)

That's it. Every client.chat.completions.create() call in your codebase now works through AIPower. No other code changes needed.

What Stays the Same

• SDK: pip install openai — same package
• API format: chat.completions.create() — identical interface
• Response format: response.choices[0].message.content — same structure
• Streaming: stream=True — works identically
• Error handling: Same exception types

What You Gain

Model Name Mapping

Your existing OpenAI model names still work. Or switch to cheaper alternatives:

# Keep using OpenAI models (with AIPower pricing)
model="openai/gpt-5.4"        # GPT-5.4 via AIPower
model="openai/gpt-4o-mini"    # GPT-4o Mini via AIPower

# Or switch to cheaper models with one parameter change
model="deepseek/deepseek-chat"   # 90% cheaper than GPT-5.4
model="qwen/qwen-turbo"          # 96% cheaper than GPT-5.4
model="zhipu/glm-4-flash"        # 99% cheaper than GPT-5.4

# Or let AI decide
model="auto"        # Best balance of quality and cost
model="auto-cheap"  # Cheapest available model

Environment Variable Approach

For production apps, use environment variables so you can switch without code changes:

import os
from openai import OpenAI

client = OpenAI(
    base_url=os.getenv("AI_BASE_URL", "https://api.aipower.me/v1"),
    api_key=os.getenv("AI_API_KEY"),
)

# .env file
# AI_BASE_URL=https://api.aipower.me/v1
# AI_API_KEY=your-aipower-key

Node.js / TypeScript Migration

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.aipower.me/v1",  // ADD
  apiKey: "your-aipower-key",             // CHANGE
});

// Everything else stays the same
const response = await client.chat.completions.create({
  model: "deepseek/deepseek-chat",
  messages: [{ role: "user", content: "Hello!" }],
});

cURL Migration

# Before (OpenAI)
curl https://api.openai.com/v1/chat/completions \
  -H "Authorization: Bearer sk-openai-key" \
  -d '{"model": "gpt-4o", "messages": [{"role": "user", "content": "Hi"}]}'

# After (AIPower) — change URL and key, that's it
curl https://api.aipower.me/v1/chat/completions \
  -H "Authorization: Bearer your-aipower-key" \
  -d '{"model": "deepseek/deepseek-chat", "messages": [{"role": "user", "content": "Hi"}]}'

FAQ

• Is it really just 2 lines? Yes. AIPower is fully OpenAI-compatible. Same SDK, same format.
• Can I still use GPT models? Yes. All OpenAI models are available through AIPower.
• What about function calling? Supported on models that support it (GPT, Claude, DeepSeek).
• Is there added latency? About 20-50ms per request. Negligible for most applications.
• Can I switch back? Yes. Change 2 lines back. No lock-in.

Try it now at aipower.me — 50 free API calls, migration takes 2 minutes. Keep your code, access 16 models, save up to 90%.