DeepSeek (V3 / R1)
LLMby DeepSeek
Chinese AI lab offering extremely cost-effective models. DeepSeek-V3 for general tasks and DeepSeek-R1 for advanced reasoning. Known for breakthrough price-performance ratio and open-source availability.
Registration & API Key Steps
Step 2
Register with email or phone number (Chinese phone numbers supported).
Step 3
Verify your account through the confirmation code.
Step 4
Navigate to API Keys page and create a new API key.
Step 5
You will automatically receive 10 RMB in free credits.
Step 6
Top up via Alipay or WeChat Pay for continued use.
Pricing
| Tier | Price | Features |
|---|---|---|
| DeepSeek-V3.2 | $0.25 / $0.38 per 1M tokens | Input / Output. Latest general model. 163K context. |
| DeepSeek-V3.1 | $0.15 / $0.75 per 1M tokens | Input / Output. 32K context. |
| DeepSeek-R1 | $0.70 / $2.50 per 1M tokens | Input / Output. Advanced reasoning model. 64K context. |
| DeepSeek-R1-0528 | $0.40 / $1.75 per 1M tokens | Input / Output. Updated reasoning model. 163K context. |
| Cache hits | 90% off input price | Automatic prompt caching for repeated prefixes. |
Application Tips
Tip 1
DeepSeek offers the best price-performance ratio in the industry — V3 is 10-50x cheaper than GPT-4o.
Tip 2
R1 reasoning model is open-source and can be self-hosted for zero API cost.
Tip 3
API is OpenAI-compatible — just change the base_url and model name.
Tip 4
Automatic prompt caching gives 90% discount on cached content.
Tip 5
Third-party providers (SiliconFlow, Volcano Engine) offer additional free credits for DeepSeek models.
Tip 6
For higher availability, use third-party hosted versions as official API can get congested during peak hours.
China Access Solutions
Access Solution
Directly accessible in China without VPN. Supports Chinese phone registration, Alipay/WeChat payment. Also available via domestic providers: SiliconFlow, Volcano Engine, Alibaba Bailian.
Code Example
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", # V3
messages=[
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Solve this step by step: What is 15% of 340?"}
]
)
print(response.choices[0].message.content)
# For reasoning model:
# model="deepseek-reasoner" # R1
# --- cURL example ---
# curl https://api.deepseek.com/chat/completions \
# -H "Authorization: Bearer your-api-key" \
# -H "Content-Type: application/json" \
# -d '{"model":"deepseek-chat","messages":[{"role":"user","content":"Hello!"}]}'Rate Limits
| Tier | Limits |
|---|---|
| Default | Default: 60 RPM, 1M TPM (V3), 500K TPM (R1). Unlimited daily requests. Can apply for higher limits. |
Recommended Use Cases
Related API Guides
OpenAI GPT-4o / GPT-4.1 / o3
OpenAI
OpenAI's flagship LLM family including GPT-4o for multimodal tasks, GPT-4.1 for long-context coding, and o3 for advanced reasoning. Industry-leading models with the largest developer ecosystem.
Anthropic Claude (Sonnet 4.5 / Opus 4.5)
Anthropic
Anthropic's Claude model family excels in nuanced reasoning, safety, and long-context tasks. Claude Sonnet 4.5 offers the best balance of cost and performance, while Opus 4.5 delivers frontier intelligence.
Google Gemini (2.5 Pro / 2.5 Flash)
Google's Gemini models offer a generous free tier, 1M token context window, and strong multimodal capabilities. Gemini 2.5 Pro leads in reasoning, while Flash models provide cost-effective alternatives.
Meta Llama 4 (Scout / Maverick)
Meta
Meta's open-source Llama 4 models are free to use and available through multiple cloud providers. Llama 4 Scout and Maverick offer competitive performance at extremely low cost through partner APIs.