Model Overview

Available Models

Pricing and Context

ModelInput PriceOutput PriceContext Length
deepseek-reasoner-all¥0.0064/K tokens¥0.0032/K tokens32K
deepseek-r1¥0.0128/K tokens¥0.0032/K tokens32K
deepseek-reasoner¥0.0128/K tokens¥0.0032/K tokens32K
deepseek-v3¥0.0032/K tokens¥0.0016/K tokens32K
deepseek-search¥0.0020/K tokens¥0.0010/K tokens32K

Model Capabilities

ModelCapabilities
deepseek-reasoner-allChat Reasoning Web File Analysis
deepseek-r1Chat Reasoning
deepseek-reasonerChat Reasoning
deepseek-v3Chat
deepseek-searchChat

Model Features

🌟

deepseek-reasoner-all

Our most powerful DeepSeek-R1 67B model with web browsing and text analysis capabilities powered by LangChain integration.

🧮

deepseek-r1

High-performance 67B model matching OpenAI-01 capabilities in mathematics, coding, and reasoning tasks. Supports 32K context window.

🤖

deepseek-reasoner

Specialized reasoning model based on DeepSeek-R1 67B, optimized for complex problem-solving and coding tasks.

🚀

deepseek-v3

Advanced model comparable to GPT-4 and Claude-3.5-Sonnet, trained on updated data surpassing LLama 3.1 and Qwen 2.5.

🔍

deepseek-search

Optimized search model based on deepseek-v3 architecture.

Model Selection Guide

  1. Choose deepseek-reasoner-all for:
    • Maximum performance requirements
    • Web browsing capabilities
    • Comprehensive text analysis
  2. Select deepseek-r1 or deepseek-reasoner for:
    • Advanced reasoning tasks
    • Complex coding projects
    • Mathematical computations
  3. Use deepseek-v3 for:
    • Cost-effective general conversations
    • Balanced performance
    • Everyday AI assistance
  4. Choose deepseek-search for:
    • Search-specific applications
    • Information retrieval tasks
    • Optimized search performance