How to Build Your Own AI API Gateway (70x Cheaper Than GPT-4o)

#ai #python #api #opensource

Why I Built This

DeepSeek's API costs ¥1 per million tokens. GPT-4o costs ¥70. That's a 70x difference.

But not everyone wants to deal with API key management, billing, and deployment. So I built a one-command gateway.

What It Does

OpenAI-compatible endpoint — drop-in replacement for any OpenAI SDK
Routes to DeepSeek — 70x cheaper, same quality for most tasks
API key management — generate and manage keys for your team
Usage tracking — see who's using what
Self-hosted — your data, your server

One Command Deploy

git clone https://github.com/matengtian/deepseek-gateway.git
cd deepseek-gateway
pip install -r requirements.txt
DEEPSEEK_API_KEY=sk-your-key python server.py

That's it. You now have an OpenAI-compatible API endpoint.

Use It Like OpenAI

import openai
client = openai.OpenAI(
    api_key="YOUR_GATEWAY_KEY",
    base_url="http://your-server:8080/v1"
)
response = client.chat.completions.create(
    model="deepseek-chat",
    messages=[{"role": "user", "content": "Hello!"}]
)