Migrate from OpenAI to AnAiApi in 3 Lines (Keep Your Code)

Switching LLM providers usually means rewriting clients, re-reading docs, and re-testing everything. With AnAiApi it's three lines: change the base URL and the key, keep the rest. Here's the exact swap — and why it's worth doing.

The 3-line change

If you already use the OpenAI Python SDK, this is the whole diff:

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_ANAI_API_KEY",          # 1. your AnAiApi key
    base_url="https://www.anaiapi.com/v1"  # 2. point at AnAiApi
)
# 3. same call shape — just pick a model
resp = client.chat.completions.create(
    model="deepseek-v4-pro",
    messages=[{"role": "user", "content": "Hello!"}]
)

That's it. Your retries, streaming, function-calling and message format all stay the same — OpenAI's client just talks to a different endpoint.

cURL

curl https://www.anaiapi.com/v1/chat/completions \
  -H "Authorization: Bearer $ANAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4-pro","messages":[{"role":"user","content":"Hello!"}]}'

Node.js

import OpenAI from "openai";
const client = new OpenAI({
  apiKey: "YOUR_ANAI_API_KEY",
  baseURL: "https://www.anaiapi.com/v1"
});

What stays the same vs. what changes

Why developers actually switch

AnAiApi is an OpenAI-compatible gateway to China's strongest models, billed in USD. Two reasons teams move:

• Cost. DeepSeek-V4-Pro is ~80% cheaper per output token than GPT-4o, and the Flash tier is ~94% cheaper — same code, far lower bill.
• Model choice. Access DeepSeek, Moonshot Kimi, Zhipu GLM and MiniMax through one integration, no separate Chinese accounts.

Models you can call today

deepseek-v4-pro deepseek-v4-flash kimi-k3 glm-4.5 MiniMax-M3

Swap the model field to route between them — or build multi-model fallback without extra integrations. The full list is in the API documentation.

FAQ

Do I need to rewrite my app?

No. Keep the OpenAI SDK; only change base_url and the API key. Message format, streaming and tool calls are unchanged.

Which model should I start with?

deepseek-v4-pro for quality, deepseek-v4-flash for cheap high-volume calls. Swap anytime.