Back to all LLMs
DeepSeek

DeepSeek-R1

Overview

DeepSeek-R1 is the first-generation reasoning model built atop DeepSeek-V3 (671B total parameters, 37B activated per token). It incorporates large-scale reinforcement learning (RL) to enhance its chain-of-thought and reasoning capabilities, delivering strong performance in math, code, and multi-step reasoning tasks.

Model Information

Max Input
131K
Tokens
Input Price
$0.55
per 1M Tokens
Output price
$2.19
per 1M Tokens
Size
671
Billion Parameters
Release date
2025-01-20
Licence
MIT License
Parameters
671B
Input Context Length
131,072
Tokens
Output Context Length
131,072
Tokens

Features & Capabilities

Web Access

Yes
Real-time access to current web information

Multimodal

Yes
Ability to process multiple data types (text, images, etc...)

API

Sample code and API for GPT-4o Search Preview
Create API Key