DeepSeek: DeepSeek V3.1

DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates.

Context
164K
Max output
33K
Input price
$0.21/1M tokens
Output price
$0.79/1M tokens

Capabilities

vision
tool call
structured output
reasoning
json mode
streaming
fine tuning
batch

Details

Model IDdeepseek/deepseek-chat-v3.1
Provider OpenRouter
Creator DeepSeek
License
Parameters
Status
Input modalitiestext
Output modalitiestext
Architecture
Knowledge cutoff
Training data cutoff
Release date
Deprecation date
Typechat
Reasoning tokens
Max input
Open weight
Sourceofficial
Last updated

Tools

Function Callingfunction_calling
Call external functions and APIs

Pricing

Input
$0.21
Output
$0.79
Cache write
Cache read
$0.13
Batch in
Batch out

Family Comparison: deepseek-chat

ModelContextMax outPricing
DeepSeek: DeepSeek V3 0324164K16K$0.2$0.77
DeepSeek: DeepSeek V3.1164K33K$0.21$0.79
DeepSeek: DeepSeek V3164K16K$0.32$0.89

API

GET/v1/models/openrouter/deepseek/deepseek-chat-v3.1