How to use DeepSeek API: 🚀 Getting Started with the DeepSeek API

📥 Model Downloads

DeepSeek models are freely available for download. You can host and run them on your own infrastructure. Below are the available versions and their resources:

---

🔹 DeepSeek-R1 Models

Model	Total Params	Activated Params	Context Length	Download
DeepSeek-R1-Zero	671B	37B	128K	🤗 HuggingFace
DeepSeek-R1	671B	37B	128K	🤗 HuggingFace

Note: Both DeepSeek-R1-Zero and DeepSeek-R1 are trained on the DeepSeek-V3-Base.
For detailed architecture information, please refer to the DeepSeek-V3 repository.

---

🔹 DeepSeek-R1-Distill Models

Model	Base Model	Download
DeepSeek-R1-Distill-Qwen-1.5B	Qwen2.5-Math-1.5B	🤗 HuggingFace
DeepSeek-R1-Distill-Qwen-7B	Qwen2.5-Math-7B	🤗 HuggingFace
DeepSeek-R1-Distill-Llama-8B	Llama-3.1-8B	🤗 HuggingFace
DeepSeek-R1-Distill-Qwen-14B	Qwen2.5-14B	🤗 HuggingFace
DeepSeek-R1-Distill-Qwen-32B	Qwen2.5-32B	🤗 HuggingFace
DeepSeek-R1-Distill-Llama-70B	Llama-3.3-70B-Instruct	🤗 HuggingFace

Note: The DeepSeek-R1-Distill models are fine-tuned versions of open-source models using output samples from DeepSeek-R1.
Some configurations and tokenizers have been modified—please use DeepSeek’s provided settings for proper deployment.

Deepseek Open Platform

At same time you can access deepseek through Deepseek open platform.

🔐 Step 1: Get an API Key

1. Go to the DeepSeek Open Platform.
2. Sign up or log in.
3. Navigate to “API Keys” from the sidebar.
4. Click “Create API Key”, name it if desired, and copy the key securely.

   ⚠️ You’ll only see the key once, so make sure to store it safely.

---

🔧 Step 2: Set Up Authentication

Use Bearer Token authentication in your HTTP headers:

• HTTP Header:

   Authorization: Bearer YOUR_API_KEY

• API Endpoint:

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

Ensure you include the Authorization header in all your API requests

🧠Step 3: Choose a Model

DeepSeek offers two primary models:​

• deepseek-chat (V3): A general-purpose model trained on extensive text data, suitable for a wide range of tasks.

• deepseek-reasoner (R1): Optimized for advanced reasoning, mathematics, and coding tasks. ​ Learn R, Python & Data Science Online

Specify your chosen model in the model parameter of your API request.​

💻 Step 4: Make an API Call (Python Example)

Here’s how to make a basic API call using Python:

url = "https://api.deepseek.com/v1/chat/completions"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}
data = {
    "model": "deepseek-chat",
    "messages": [{"role": "user", "content": "Hello, DeepSeek!"}]
}

response = requests.post(url, headers=headers, json=data)
print(response.json())

Replace YOUR_API_KEY with your actual API key.

🧪 Step 5: Test with Tools like Apidog

For testing and debugging, you can use tools like Apidog:​

1. Install Apidog and create a new API request.

2. Set the HTTP method to POST.

3. Enter the API endpoint: https://api.deepseek.com/v1/chat/completions.

4. Configure headers:

   1. Authorization: Bearer YOUR_API_KEY

   2. Content-Type: application/json

5. Add the request body in JSON format:

{
  "model": "deepseek-chat",
  "messages": [{"role": "user", "content": "Hello, DeepSeek!"}]
}

6. Send the request and observe the response.

📚 Additional Resources

Here are some useful links to learn more about DeepSeek and its ecosystem:

• 🔗 DeepSeek GitHub Repository
  Access source code, model weights, and documentation for DeepSeek models.

• 🔗 DeepSeek-R1 Evaluation
  Detailed benchmarks comparing DeepSeek-R1 with other LLMs like GPT-4o.

• 🔗 DeepSeek V2 & V3 Models
  Explore previous versions of DeepSeek, including improvements in architecture and capabilities.

• 🔗 DeepSeek Open Platform
  Sign up to generate API keys, manage usage, and access model endpoints.

• 🔗 HuggingFace DeepSeek Models
  Download models for offline or on-premise use.

• 🔗 DeepSeek on X (Twitter)
  Follow for updates, community news, and release announcements.