Skip to main content
DEDeepSeek/ DeepSeek

DeepSeek V3.2

deepseek/deepseek-v3.2

DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...

Model information

Released
2025-12-01
Context window
164K
Max output
66K
Model parameters
685B
Weights
Open

Modalities

Input
Text
Output
Text

Identity

Version
deepseek-v3.2
License
MIT

Resources

Capabilities

Open weightReasoningStructuredTemperatureTool call

Providers

ProviderProvider model IDProtocolContextMax outputInputOutputCapabilitiesNotesPricingStreaming
OpenRouterdeepseek/deepseek-v3.2openai_chat164K66K
ReasoningTool callStructuredTemperature
Reasoning disabled by default, Hugging Face: deepseek-ai/DeepSeek-V3.2
input $0.269/M · output $0.4/M
cache read $0.135/M
Yes
NanoGPTdeepseek/deepseek-v3.2openai_chat164K66K
AttachmentsTool callStructured
-
input $0.28/M · output $0.42/M
cache read $0.14/M
Yes
ZenMuxdeepseek/deepseek-v3.2openai_chat164K66K
ReasoningTool callTemperature
-
input $0.28/M · output $0.43/M
Yes
TensorXdeepseek/deepseek-v3.2openai_chat164K66K
ReasoningTool callTemperature
-
input $0.3/M · output $0.5/M
cache read $0.075/M · cache write $0.375/M
Yes
Kilo Gatewaydeepseek/deepseek-v3.2openai_chat164K66K
ReasoningTool callStructuredTemperature
-
input $0.269/M · output $0.4/M
cache read $0.135/M
Yes
Vercel AI Gatewaydeepseek/deepseek-v3.2openai_chat164K66K
Temperature
-
input $0.28/M · output $0.42/M
cache read $0.028/M
Yes
NovitaAIdeepseek/deepseek-v3.2openai_chat164K66K
ReasoningTool callStructuredTemperature
-
input $0.269/M · output $0.4/M
cache read $0.135/M
Yes
DeepSeekdeepseek-v3.2custom164K66K
---Yes

FAQ