How to Use DeepSeek V4 with OpenAI Python SDK

July 7, 2026 — 6 min read tutorial openai-sdk

One of the best things about DeepSeek V4 is that it's fully compatible with the OpenAI API format. If you already have code written for OpenAI, switching to DeepSeek is as simple as changing two lines.

In this guide, we'll walk through the complete setup: getting your API key, configuring the SDK, and making your first request — all in under 5 minutes.

What You'll Need

• A PayPal account (to purchase API credits)
• Python 3.8+ with the openai package installed
• 2 minutes of your time

Step 1: Get Your API Key

Head over to Token Gateway, choose a plan, and pay with PayPal. Your API key will be generated instantly and displayed on your dashboard.

That's it. No phone verification. No ID checks. No waiting for approval.

Step 2: Install the OpenAI SDK

You probably already have this, but if not:

pip install openai

Step 3: Make Your First API Call

The only changes needed are the base_url and api_key:

from openai import OpenAI

client = OpenAI(
    api_key="tg-your-api-key-here",
    base_url="https://token.mall199.com/v1"
)

response = client.chat.completions.create(
    model="deepseek-flash",  # or "deepseek-pro"
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Write a Python function to calculate Fibonacci numbers."}
    ]
)

print(response.choices[0].message.content)

Step 4: Streaming Responses

Streaming works exactly like OpenAI. Just add stream=True:

stream = client.chat.completions.create(
    model="deepseek-flash",
    messages=[{"role": "user", "content": "Tell me a story"}],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

Using cURL

No SDK? No problem. Raw HTTP works too:

curl https://token.mall199.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer tg-your-api-key-here" \
  -d '{
    "model": "deepseek-flash",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Using Node.js

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "tg-your-api-key-here",
  baseURL: "https://token.mall199.com/v1",
});

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

console.log(response.choices[0].message.content);

Integrating with LangChain

DeepSeek works seamlessly with LangChain through the OpenAI chat model wrapper:

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="deepseek-flash",
    openai_api_key="tg-your-api-key-here",
    openai_api_base="https://token.mall199.com/v1"
)

response = llm.invoke("Explain quantum computing in simple terms")
print(response.content)

Model Options

Tips

• Start with deepseek-flash for most use cases — it's incredibly capable for its price.
• Switch to deepseek-pro for complex reasoning tasks.
• Monitor your usage from the dashboard to avoid surprises.
• All packages include generous rate limits — no hidden throttling.