DEV Community

Cover image for DeepSeek API: Free Access to R1 Reasoning and V3 Chat Models
toolfreebie
toolfreebie

Posted on • Originally published at toolfreebie.com

DeepSeek API: Free Access to R1 Reasoning and V3 Chat Models

#ai #api #opensource

Why DeepSeek API Is the Best Budget-Friendly Free AI API

If you want powerful AI at the lowest possible cost — or completely free — DeepSeek API is hard to beat. DeepSeek offers two world-class models: DeepSeek-V3 (a 671B MoE chat model) and DeepSeek-R1 (a frontier reasoning model that rivals OpenAI o1). New users get free trial credits on sign-up, and even the paid tier is priced up to 20× cheaper than GPT-4o. Best of all, it’s fully OpenAI SDK-compatible — no code changes needed.

In this guide, you’ll learn how to get your free API key, call the API with Python, use the R1 reasoning model, integrate DeepSeek into OpenClaw for a free AI agent, and see how it stacks up against other free APIs.

DeepSeek Models Overview

Model API Name Context Window Best For
DeepSeek-V3 deepseek-chat 64K tokens General tasks, coding, summarization, chat
DeepSeek-R1 deepseek-reasoner 64K tokens Math, logic, complex reasoning, step-by-step thinking

DeepSeek-V3 is one of the strongest open-weight models available — it outperforms GPT-4o on many coding and reasoning benchmarks. DeepSeek-R1 is a dedicated reasoning model with built-in chain-of-thought, matching OpenAI o1’s performance at a fraction of the cost.

Free Tier & Pricing

Details
Free Trial Credits New accounts receive free credits on registration (no credit card required)
DeepSeek-V3 (Input) ~$0.27 / 1M tokens (cache miss), ~$0.07 / 1M tokens (cache hit)
DeepSeek-V3 (Output) ~$1.10 / 1M tokens
DeepSeek-R1 (Input) ~$0.55 / 1M tokens (cache miss)
DeepSeek-R1 (Output) ~$2.19 / 1M tokens
Credit Card Required No (for free trial)

To put the pricing in perspective: running 1 million tokens through DeepSeek-V3 costs about $0.27, compared to $5.00 for GPT-4o input. For most hobby projects and prototypes, the free trial credits will cover weeks of usage.

How to Get Your Free API Key

  1. Go to platform.deepseek.com and sign up with your email
  2. Complete email verification — no credit card required
  3. Navigate to API Keys in the left sidebar
  4. Click “Create new API key”, give it a name, and copy it
  5. Your free trial credits are automatically applied to your account

Using the API with Python

Install the OpenAI SDK

DeepSeek’s API is fully compatible with the OpenAI SDK. No separate package needed.

pip install openai
Enter fullscreen mode Exit fullscreen mode

Basic Chat with DeepSeek-V3 

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_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 coding assistant."},
        {"role": "user", "content": "Write a Python function to find all prime numbers up to N using the Sieve of Eratosthenes"}
    ]
)

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

Using DeepSeek-R1 for Reasoning Tasks

DeepSeek-R1 is a thinking model. It generates a chain-of-thought internally before producing its final answer. You can access the reasoning trace via reasoning_content.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_DEEPSEEK_API_KEY",
    base_url="https://api.deepseek.com"
)

response = client.chat.completions.create(
    model="deepseek-reasoner",
    messages=[
        {"role": "user", "content": "A train travels 120km in 1.5 hours, then 80km in 45 minutes. What is the average speed for the whole journey?"}
    ]
)

# The reasoning trace (chain-of-thought)
reasoning = response.choices[0].message.reasoning_content
print("Thinking:", reasoning)

# The final answer
answer = response.choices[0].message.content
print("Answer:", answer)
Enter fullscreen mode Exit fullscreen mode

Streaming Responses 

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_DEEPSEEK_API_KEY",
    base_url="https://api.deepseek.com"
)

stream = client.chat.completions.create(
    model="deepseek-chat",
    messages=[{"role": "user", "content": "Explain how neural networks learn"}],
    stream=True
)

for chunk in stream:
    delta = chunk.choices[0].delta
    if delta.content:
        print(delta.content, end="", flush=True)
Enter fullscreen mode Exit fullscreen mode

Function Calling

DeepSeek-V3 supports tool/function calling, making it ideal for building agents.

from openai import OpenAI
import json

client = OpenAI(
    api_key="YOUR_DEEPSEEK_API_KEY",
    base_url="https://api.deepseek.com"
)

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get the current weather for a location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {"type": "string", "description": "City name"}
                },
                "required": ["location"]
            }
        }
    }
]

response = client.chat.completions.create(
    model="deepseek-chat",
    messages=[{"role": "user", "content": "What's the weather in Tokyo?"}],
    tools=tools,
    tool_choice="auto"
)

tool_call = response.choices[0].message.tool_calls[0]
print("Function called:", tool_call.function.name)
print("Arguments:", tool_call.function.arguments)
Enter fullscreen mode Exit fullscreen mode

Connect DeepSeek to OpenClaw (Free AI Agent)

Pair DeepSeek’s free API with OpenClaw to build a fully autonomous AI agent that can search the web, run code, manage files, and communicate on WhatsApp or Telegram — all for free.

Quick Setup 

npm install -g openclaw@latest
openclaw onboard
Enter fullscreen mode Exit fullscreen mode

When prompted for the provider, select OpenAI-compatible and enter https://api.deepseek.com as the base URL with your DeepSeek API key.

Manual Configuration

Edit ~/.openclaw/openclaw.json:

{
  "models": {
    "mode": "merge",
    "providers": {
      "deepseek": {
        "baseUrl": "https://api.deepseek.com",
        "apiKey": "YOUR_DEEPSEEK_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "deepseek-chat",
            "name": "DeepSeek V3",
            "reasoning": false,
            "input": ["text"],
            "contextWindow": 65536,
            "maxTokens": 8192
          },
          {
            "id": "deepseek-reasoner",
            "name": "DeepSeek R1",
            "reasoning": true,
            "input": ["text"],
            "contextWindow": 65536,
            "maxTokens": 8192
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "deepseek/deepseek-chat"
      },
      "models": {
        "deepseek/deepseek-chat": {}
      }
    }
  }
}
Enter fullscreen mode Exit fullscreen mode

Use deepseek-chat for everyday agent tasks (fast and cheap) and switch to deepseek-reasoner when the agent needs to solve complex multi-step problems.

DeepSeek vs Other Free AI APIs

Feature DeepSeek Google Gemini Groq Alibaba Bailian
Best Model DeepSeek-R1 (reasoning) Gemini 2.5 Pro Llama 3.3 70B Qwen 3.6-Plus
Free Tier Trial credits on signup 100-1,000 RPD 14,400 RPD 1M tokens/model
Paid Pricing Cheapest (~$0.27/1M) Moderate Very low Very low
Reasoning Model Yes (R1) Gemini 2.5 Pro (thinking) No QwQ (limited)
Context Window 64K 1M 128K 1M
OpenAI Compatible Yes Yes Yes Yes
Credit Card No No No No
Speed Moderate Fast Ultra-fast Fast
Multimodal No Yes No Yes

When to Use DeepSeek API

  • Coding & debugging — DeepSeek-V3 excels at writing, reviewing, and fixing code
  • Math & logic problems — DeepSeek-R1’s chain-of-thought reasoning handles complex problems step by step
  • Cost-sensitive production apps — At ~$0.27/1M input tokens, V3 is one of the most affordable APIs for high-volume use
  • OpenAI migration — Full drop-in replacement, no code rewrite needed
  • Agentic workflows — Function calling support makes it suitable for tool-using agents

Important Limitations

  • No multimodal support — DeepSeek API is text-only; use Gemini if you need image/audio/video input
  • 64K context window — For very long documents, consider Gemini (1M tokens) or Alibaba Bailian
  • Free credits are limited — Once trial credits run out, you need to top up (though pricing is very low)
  • R1 streaming quirk — When streaming R1, reasoning_content arrives before the final answer; handle both in your stream loop

Related Reads

Final Thoughts

DeepSeek API is the best choice if you need a powerful reasoning model at the lowest cost. DeepSeek-R1 is genuinely competitive with OpenAI o1 for math, coding, and logical reasoning — and it costs a fraction of the price. DeepSeek-V3 is an excellent everyday chat and coding model that’s fully OpenAI-compatible.

For pure speed, Groq is faster. For multimodal and long-context tasks, Gemini has the edge. But for the best combination of reasoning capability and cost-efficiency, DeepSeek is the clear winner among free AI APIs in 2026.

Get started: platform.deepseek.com

Originally published at toolfreebie.com.

Top comments (0)