DeepYard
D

DeepSeek V3.2

8

DeepSeek-V3.2 is a large language model designed to harmonize high computational...

Open Sourcepaid

About

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 that reduces training and inference cost while preserving quality in long-context scenarios. A scalable reinforcement learning post-training framework further improves reasoning, with reported performance in the GPT-5 class, and the model has demonstrated gold-me

Details

Modalitytext
Context Window163.8K tokens
Release DateDec 1, 2025
API AvailableYes
Hostingself-hosted, api
Output Speed140 tokens/sec
Time to First Token500ms
Quality Index86/100
Coding Index85/100
Reasoning Index82/100

Benchmarks

Chatbot Arena ELO
1435/1500
MMLU-Pro
85/100
SWE-bench Verified
77.2/100
MATH-500
88.5/100
GPQA Diamond
71.2/100
HumanEval
93/100
Artificial Analysis Index
38/100

Tags

open-weightapichat