How to beginner · 3 min read

How to use DeepSeek R1 for complex reasoning

Q: How to use DeepSeek R1 for complex reasoning

Use the deepseek-reasoner model via the OpenAI-compatible DeepSeek API by sending your reasoning prompts to client.chat.completions.create(). This model specializes in complex reasoning and can be accessed with your API key and the correct base URL.

Quick answer

Use the deepseek-reasoner model via the OpenAI-compatible DeepSeek API by sending your reasoning prompts to client.chat.completions.create(). This model specializes in complex reasoning and can be accessed with your API key and the correct base URL.

PREREQUISITES

Python 3.8+
DeepSeek API key
pip install openai>=1.0

Setup

Install the openai Python package (version 1.0 or higher) and set your DeepSeek API key as an environment variable. Use the DeepSeek API endpoint https://api.deepseek.com to access their models.

bash

pip install openai>=1.0

Step by step

This example demonstrates how to call the deepseek-reasoner model for complex reasoning tasks using the OpenAI-compatible DeepSeek API client.

python

import os
from openai import OpenAI

# Initialize DeepSeek client with base_url
client = OpenAI(api_key=os.environ["DEEPSEEK_API_KEY"], base_url="https://api.deepseek.com")

# Define the reasoning prompt
messages = [
    {"role": "user", "content": "Explain the steps to solve a complex math problem involving calculus and linear algebra."}
]

# Create chat completion request
response = client.chat.completions.create(
    model="deepseek-reasoner",
    messages=messages,
    max_tokens=1024
)

# Extract and print the reasoning output
print(response.choices[0].message.content)

output

Explain the steps to solve a complex math problem involving calculus and linear algebra.

1. Understand the problem statement and identify the variables.
2. Break down the problem into calculus and linear algebra components.
3. Apply differentiation or integration techniques as needed.
4. Use matrix operations or vector spaces for the linear algebra part.
5. Combine results to reach the final solution.
6. Verify the solution by checking boundary conditions and consistency.

Common variations

You can adjust max_tokens for longer or shorter responses, switch to other DeepSeek models like deepseek-chat for general chat, or implement asynchronous calls using Python asyncio with the OpenAI SDK.

python

import asyncio
from openai import OpenAI

async def async_reasoning():
    client = OpenAI(api_key=os.environ["DEEPSEEK_API_KEY"], base_url="https://api.deepseek.com")
    response = await client.chat.completions.acreate(
        model="deepseek-reasoner",
        messages=[{"role": "user", "content": "Explain the theory of relativity in simple terms."}],
        max_tokens=512
    )
    print(response.choices[0].message.content)

asyncio.run(async_reasoning())

output

The theory of relativity, developed by Einstein, explains how space and time are linked and how gravity affects them. It shows that time can slow down or speed up depending on how fast you move or how strong gravity is.

Troubleshooting

If you receive authentication errors, verify your DEEPSEEK_API_KEY environment variable is set correctly.
For timeout or connection issues, check your network and ensure the base URL https://api.deepseek.com is reachable.
If responses are incomplete, increase max_tokens or check for rate limits on your API key.

✅

Key Takeaways

Use the OpenAI-compatible SDK with base_url="https://api.deepseek.com" to access DeepSeek models.
The deepseek-reasoner model is optimized for complex reasoning tasks.
Adjust max_tokens and use async calls for flexible integration.
Always set your API key in the environment variable DEEPSEEK_API_KEY.
Check network connectivity and API limits if you encounter errors.

Verified 2026-04 · deepseek-reasoner, deepseek-chat

Verify ↗