Which is better, DeepSeek-V3.2 or Qwen3.5 2B?

DeepSeek-V3.2 vs Qwen3.5 2B

23 SHARED BENCHMARKS

Across 23 shared benchmarks, DeepSeek-V3.2 scores higher on 22 and Qwen3.5 2B on 1. The widest gap is Terminal-Bench Hard, where DeepSeek-V3.2 scores 35.6 against 3.8.

DEEPSEEKVSALIBABA23 SHARED22–1 HEAD-TO-HEAD

BenchmarkDeepSeek-V3.2MARGINQwen3.5 2B

AA Agentic Index39.80.8

AA Intelligence21.56.9

AA-LCR73.325.6

AA-Omniscience-22.5-60.2

AIME 202593.119

Artificial Analysis Coding Index44.22.9

critpt2.90

gdpval18.80

GPQA Diamond8451.6

HLE40.85

IFBench6141.3

longbench_v259.838.7

mmlu_redux93.779.6

MMLU-Pro8666.5

OmniScience Accuracy338.1

OmniScience Non-Hallucination17.325.8

scicode397.2

t2-bench80.348.8

TauBench V3 - Banking18.82.5

Terminal-Bench 2.146.83

Terminal-Bench Hard35.63.8

τ²-Bench Telecom (AA run)90.681.6

τ³-Bench69.26.2

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.

Compare DeepSeek-V3.2 withALL PAIRINGS →

DeepSeekDeepSeek-V4-Pro DeepSeek-V3 DeepSeek-V4-Flash

Across the fieldQwen3.6 27B Qwen3.7 Max Qwen3.8 Max Preview Qwen3.7 Plus Preview Claude Opus 4 Qwen3.5 4B Gemini 3.6 Flash Nemotron 3 Super 120B A12B

Compare Qwen3.5 2B withALL PAIRINGS →

AlibabaQwen3.5 35B A3B Qwen3 VL 4B (Reasoning)Qwen3.5 397B A17B Qwen2.5 Instruct 72B Qwen3.5 27B Qwen3.6 Plus

Across the fieldDeepSeek-V4-Flash Claude Opus 4.7 Gemini 2.5 Pro Gemini 1.5 Pro Claude Sonnet 4 Claude Sonnet 5 Muse Spark 1.1 GLM-5.1