Deepseek
Date
Price

deepseek-v4-pro

The latest flagship AI model released by the DeepSeek series represents the current highest standard in both scale and performance among open-source models.
Model
LLM
Model capability: thinkingModel capability: function_call
Input:
$0.43/1M tokensstarting from
Output:
$0.86/1M tokensstarting from

deepseek-v4-flash

DeepSeek’s newly released language model, designed for high-performance production scenarios, is specifically optimized for ultimate inference efficiency and response speed.
Model
LLM
Model capability: thinkingModel capability: function_call
Input:
$0.14/1M tokensstarting from
Output:
$0.28/1M tokensstarting from

deepseek-v3.2-thinking

The deep reasoning mode optimized based on DeepSeek-v3.2 excels at multi-turn reasoning and tool calls to provide more detailed answers.
Model
LLM
Model capability: thinkingModel capability: function_call
Input:
$0.29/1M tokensstarting from
Output:
$0.43/1M tokensstarting from

deepseek-v3.2

A “daily workhorse” reasoning model that balances reasoning ability and output efficiency.
Model
LLM
Model capability: function_call
Input:
$0.29/1M tokensstarting from
Output:
$0.43/1M tokensstarting from

deepseek-v3.2-exp

Experimental version of DeepSeek-V3.2
Model
LLM
Model capability: function_call
Input:
$0.29/1M tokensstarting from
Output:
$0.43/1M tokensstarting from

deepseek-v3.1

Deepseek Dual-Mode Inference Model
Model
LLM
Model capability: function_call
Input:
$0.286/1M tokensstarting from
Output:
$1.15/1M tokensstarting from

deepseek-vl2

Open-source multimodal large model with a Mixture of Experts (MoE) architecture
Model
LLM
Model capability: image
Input:
$0.165/1M tokens
Output:
$0.165/1M tokens

deepseek-r1

A model optimized for complex reasoning tasks
Model
LLM
Model capability: function_call
Input:
$0.6/1M tokens
Output:
$2.3/1M tokens