Getting Started with DeepSeek API: A Quick Start Guide

DeepSeek API offers a powerful and versatile platform for developers looking to integrate advanced AI capabilities into their applications. This article will guide you through your first API call, explaining the fundamentals and providing practical examples to get you up and running quickly.

DeepSeek API is designed to be compatible with the OpenAI API format. This compatibility allows you to seamlessly use the OpenAI SDK or other software that supports the OpenAI API by simply modifying the configuration.

DeepSeek API: Key Features

• OpenAI Compatibility: Easy integration with existing OpenAI-based tools and libraries.
• Latest Models: Access to cutting-edge models like DeepSeek-V3 (via deepseek-chat) and DeepSeek-R1 (via deepseek-reasoner).
• Flexible Output: Supports both non-streaming (default) and streaming outputs.
• Comprehensive Documentation: Detailed API documentation and guides available.

Prerequisites

Before making your first API call, ensure you have the following:

• An API key from the DeepSeek Platform. This is essential for authenticating your requests.
• The OpenAI SDK installed (if using Python or Node.js examples). You can install it using pip3 install openai (Python) or npm install openai (Node.js).

Configuration

To connect to the DeepSeek API, you'll primarily need the base_url and your api_key. Here's how they are used:

For OpenAI compatibility, you can also set the base_url to https://api.deepseek.com/v1. Note that the v1 here is for compatibility and doesn't relate to model versions.

Making Your First API Call: A Step-by-Step Guide

Let's walk through a basic example using the deepseek-chat model (now upgraded to DeepSeek-V3) to generate a conversational response.

Using cURL

curl https://api.deepseek.com/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <DeepSeek API Key>" \
  -d '{
  "model": "deepseek-chat",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
  ],
  "stream": false
}'

• Replace <DeepSeek API Key> with your actual API key.
• This command sends a POST request to the DeepSeek API.
• The messages array contains a system message defining the assistant's role and a user message with the initial prompt.
• stream: false specifies a non-streaming output.

Using Python

from openai import OpenAI

client = OpenAI(
    api_key="<DeepSeek API Key>",
    base_url="https://api.deepseek.com"
)

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)

• Remember to replace <DeepSeek API Key> with your API key.
• This script uses the OpenAI Python library to interact with the DeepSeek API.
• The chat.completions.create method sends the request and returns the response.

Using Node.js

import OpenAI from "openai";

const openai = new OpenAI({
  baseURL: 'https://api.deepseek.com',
  apiKey: '<DeepSeek API Key>'
});

async function main() {
  const completion = await openai.chat.completions.create({
    messages: [{ role: "system", content: "You are a helpful assistant." }],
    model: "deepseek-chat",
  });
  console.log(completion.choices[0].message.content);
}

main();

• As with the other examples, replace <DeepSeek API Key> with your actual API key.
• This code utilizes the OpenAI Node.js library to communicate with the DeepSeek API asynchronously.

Exploring DeepSeek's Reasoning Model

DeepSeek also offers a powerful reasoning model, DeepSeek-R1, accessible via the deepseek-reasoner model name. This reasoning model is designed for more complex tasks requiring logical inference and problem-solving.

To use DeepSeek-R1, simply replace "model": "deepseek-chat" with "model": "deepseek-reasoner" in your API calls.

Next Steps

Congratulations on making your first API call to DeepSeek! Now that you've grasped the basics, here are some suggestions for next steps:

• Explore the Models: Experiment with different models, such as DeepSeek-V3 and DeepSeek-R1, to understand their strengths.
• Parameter Tuning: Learn about parameters like temperature and token limits to fine-tune the output. See the Temperature Settings guide.
• Advanced Features: Explore advanced features such as Function Calling and JSON Output.
• Refer to Documentation: Read the comprehensive DeepSeek API documentation for in-depth information.
• Stay Updated: Keep an eye on the Updates page for the latest features and improvements.

By following this guide and exploring the available resources, you'll be well-equipped to leverage the power of DeepSeek API in your projects.