Skip to main content
DeepSeek/ DeepSeek

DeepSeek V3.1

deepseek/deepseek/chat-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. It extends the DeepSeek-V3 base with a two-phase long-context...

Released
2025-08-21
Context window
164K
Max output
33K
Weights
Open weight

Modalities

Input
Text
Output
Text

Identity

Version
deepseek-chat-v3.1
Knowledge cutoff
2025-03-31

Resources

Providers

ProviderProvider model IDProtocolContextMax outputInputOutputCapabilitiesNotesPricingStreaming
OpenRouterdeepseek/deepseek-chat-v3.1openai_chat164K33K
Text
Text
ReasoningTool callStructuredTemperatureOpen weight
Hugging Face: deepseek-ai/DeepSeek-V3.1
input $0.25/M · cache read $0.13/M
output $0.95/M
Yes
TensorXdeepseek/deepseek-chat-v3.1openai_chat164K33K
Text
Text
ReasoningTool callStructuredTemperatureOpen weight
-
input $0.2/M · cache read $0.05/M
output $0.8/M · cache write $0.25/M
Yes
Kilo Gatewaydeepseek/deepseek-chat-v3.1openai_chat164K33K
Text
Text
ReasoningTool callStructuredTemperatureOpen weight
-
input $0.27/M · cache read $0.135/M
output $1/M
Yes

Capabilities

ReasoningTool callStructuredTemperatureOpen weight

FAQ