How to Test DeepSeek Chat API in Postman (Based on Your Python Code)

When working with language models like DeepSeek or OpenAI-compatible APIs in your Python code, it’s often useful to test requests manually using Postman. This guide shows you how to replicate your Python OpenAI SDK call using raw HTTP requests in Postman.

🧠 The Goal

Your Python code does the following:

import os from dotenv import load_dotenv from openai import OpenAI # Load .env variables load_dotenv() # Read values from environment api_key = os.getenv("OPENAI_API_KEY") base_url = os.getenv("OPENAI_BASE_URL") # Create OpenAI client client = OpenAI(api_key=api_key, base_url=base_url) # Send chat completion request response = client.chat.completions.create( model="deepseek-chat", messages=[ {"role": "system", "content": "You are a helpful assistant"}, {"role": "user", "content": "Hello"}, ], stream=False ) print(response.choices[0].message.content) 

Now let’s test the same call in Postman.

✅ Step 1: Set the URL and HTTP Method

• Method: POST
• URL:

 https://api.deepseek.com/v1/chat/completions 

This matches the base_url + /v1/chat/completions endpoint used by the OpenAI-compatible API.

✅ Step 2: Set Headers

In Postman's Headers tab, add the following key-value pairs:

🔐 Replace YOUR_API_KEY with your actual API key from your .env file.

✅ Step 3: Configure the JSON Body

Go to the Body tab → Select raw → Set type to JSON.

Paste this JSON body:

{ "model": "deepseek-chat", "messages": [ { "role": "system", "content": "You are a helpful assistant" }, { "role": "user", "content": "Hello" } ], "stream": false } 

This mirrors your Python code’s messages and model.

✅ Step 4: Send the Request

Click the Send button. You should get a response like:

{ "id": "7e1894a2-12d7-4649-864a-da8984118be3", "object": "chat.completion", "created": 1752505878, "model": "deepseek-chat", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "Hello! How can I assist you today? 😊" }, "logprobs": null, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 9, "completion_tokens": 11, "total_tokens": 20, "prompt_tokens_details": { "cached_tokens": 0 }, "prompt_cache_hit_tokens": 0, "prompt_cache_miss_tokens": 9 }, "system_fingerprint": "fp_8802369eaa_prod0623_fp8_kvcache" } 

🎉 That means it’s working just like your Python script!

🛠 Bonus: Use Postman Environment Variables

To avoid hardcoding sensitive values:

1. In Postman → Environments, create a new environment.

2. Add:

• API_BASE_URL: https://api.deepseek.com
• API_KEY: sk-your-key

1. In your request:

• URL: {{API_BASE_URL}}/v1/chat/completions
• Header: Authorization: Bearer {{API_KEY}}

Now your requests are portable and secure.

📌 Summary

🔐 Don't Forget

• Never commit your .env file to version control.
• Avoid sharing Postman collections that contain raw keys.