// COMPARISON

DeepSeek-V3.2

A head-to-head benchmark comparison of DeepSeek-V3.2 across 18 evaluations.

The models

DE

DeepSeek-V3.2

DeepSeek
Language Proprietary
Params
β€”
Context
β€”
Released
Dec 2025

DeepSeek V3.2 is the official successor to V3.2-Exp and DeepSeek's flagship reasoning-first model built for agents. It outperforms GPT-5 on reasoning tasks and .

Full DeepSeek-V3.2 specs β†’

Benchmark comparison

Benchmark DeepSeek-V3.2
agentic
BrowseComp 51.4%
BrowseComp-Zh 65
MCP-Mark 38
MCP-Universe 45.9
Terminal-Bench 2.0 46.4%
Tool-Decathlon 35.2
τ²-Bench (TAU2) 80.3
code
Codeforces Elo 2,386
LiveCodeBench 83.3%
SWE-bench Verified 73.1%
coding
SWE-Bench Multilingual 70.2
general
MMLU-Pro 85%
math
AIME 2025 93.1%
HMMT February 2025 92.5%
HMMT November 2025 90.2
IMOAnswerBench 78.3
reasoning
GPQA Diamond 82.4%
Humanity's Last Exam 40.8%

Best result per row highlighted in cyan. Each benchmark links to its definition and sources; each model links to its full scorecard.

Frequently asked questions

Which of these models is best for coding? β–Ά

DeepSeek-V3.2 has the strongest result on Codeforces Elo among the models compared here. See the code-category rows in the table for the full picture.

How many benchmarks are compared? β–Ά

This comparison covers 18 benchmarks on which at least one of the selected models has a published score.

Where do the benchmark scores come from? β–Ά

Scores are aggregated from official model cards, technical reports and standard public evaluations, and link back to each benchmark's source. They are updated as new results are published.

Explore further

β†’ DeepSeek-V3.2 β†’ AI benchmarks leaderboard β†’ Build a different comparison
Build a different comparison

Models (0 selected)

Benchmarks (auto)

Select at least two models.