DEV Community

Jinze Wang
Jinze Wang

Posted on

Use DeepSeek with the OpenAI Python SDK by Changing One URL

#python #ai #apigateway #deepseek

If your application already uses the OpenAI Python SDK, testing another model does not have to mean rewriting the integration. An OpenAI-compatible endpoint lets you keep the familiar client, messages format, and response handling.

This tutorial shows a minimal DeepSeek request through JinzeAI's public beta endpoint.

1. Install the SDK 

python -m pip install openai
Enter fullscreen mode Exit fullscreen mode

2. Store the API key safely

Do not hard-code an API key or commit it to Git.

On macOS or Linux:

export JINZEAI_API_KEY="your_api_key_here"
Enter fullscreen mode Exit fullscreen mode

On PowerShell:

$env:JINZEAI_API_KEY="your_api_key_here"
Enter fullscreen mode Exit fullscreen mode

3. Create the client 

import os

from openai import OpenAI

client = OpenAI(
    base_url="https://jinzeai.cc/v1",
    api_key=os.environ["JINZEAI_API_KEY"],
)
Enter fullscreen mode Exit fullscreen mode

The important change is base_url. The rest of the SDK remains familiar.

4. Send a request 

response = client.chat.completions.create(
    model="deepseek-chat",
    messages=[
        {
            "role": "user",
            "content": "Explain API gateways in one sentence.",
        }
    ],
)

print(response.choices[0].message.content)
Enter fullscreen mode Exit fullscreen mode

Common errors

HTTP 401

Confirm that the API key is active and that your client sends it as a bearer token. If a key was ever exposed publicly, rotate it instead of continuing to use it.

HTTP 404

Confirm that the base URL is exactly:

https://jinzeai.cc/v1
Enter fullscreen mode Exit fullscreen mode

Chat requests must use POST /chat/completions.

Slow first test

Start with a short prompt and a new conversation. Large message histories can add substantial input tokens and latency.

Try the public beta

JinzeAI is inviting early beta testers outside mainland China. The beta may include a limited free test credit, and no payment is required to test it.

Model availability and usage limits may change during beta. I would especially value feedback about setup clarity, latency, and SDK compatibility.

Top comments (0)