DeepSeek R1 API
Power your applications with DeepSeek R1 API — advanced reasoning, math, and code generation at scale.
DeepSeek AI
DeepSeek R1
DeepSeek-R1 uses a MoE architecture with 671B total parameters, 37B activated parameters. It delivers high performance in reasoning, math, and code tasks, outperforming many open-source models and competing with top proprietary models, while offering scalable open-source options.
RL-Driven Reasoning
Model uses reinforcement learning (RL) without supervised fine-tuning to develop advanced reasoning patterns, including self-verification and long CoT generation.
Cold-Start Data Incorporation
DeepSeek R1 improves model stability and reduces issues like endless repetition and language mixing by integrating cold-start data before RL stages.
Model Distillation
DeepSeek-R1 transfers reasoning capabilities from larger models into smaller ones, achieving strong performance in more compact architectures.
Breakthrough capabilities
DeepSeek R1 Surpasses OpenAI and Claude
DeepSeek-R1's mixture-of-experts architecture drives competitive scalability and cost-efficiency, making it a leader among open-source AI models.
.png){kind=small}
Versus OpenAI o1 (1217)
DeepSeek-R1 outperforms OpenAI-o1-1217 in math and reasoning tasks like DROP and LiveCodeBench, showing its strength in complex problem-solving with RL-enhanced reasoning capabilities.
Versus Claude 3.5 Sonnet (1022)
DeepSeek-R1 excels in areas requiring deep mathematical reasoning and large-scale code completion as well as MMLU and Chinese comparing to Claude-3.5.
Versus GPT-4o (0513)
DeepSeek-R1 outperforms GPT-4o in reasoning and math tasks, excelling in MATH-500 and AIME 2024, and handling advanced logic and computation at scale.
Developing with
DeepSeek R1 API
Gain instant access to DeepSeek R1 through production API: AI/ML API provides direct connectivity to DeepSeek R1.
Read more: Technical Model Card.
Try Playground.
100+ AI Models
Cost-Effective API
High Performance
Serverless Architecture
import os
from openai import OpenAI
client = OpenAI(
base_url="<YOUR_API_KEY>",
api_key="https://api.aimlapi.com/v1",
)
response = client.chat.completions.create(
model="deepseek/deepseek-r1",
messages=[
{
"role": "system",
"content": "You are an AI assistant who knows everything.",
},
{
"role": "user",
"content": "Tell me, why is the sky blue?"
},
],
)
message = response.choices[0].message.content
print(f"Assistant: {message}")How to Get a Deepseek R1 API Key with AI/ML API
Sign Up
Begin by signing up on our AI/ML API platform. Create your account to gain access to a wide range of powerful AI models and tools.
Create API Key
In the Playground, navigate to the Key Management section and click on Create API Key. You can easily activate or deactivate your keys as needed.
Integrate AI
After creating your API key, you can integrate AI models into your application by following the guidelines provided in our API reference.
