Which is better, DeepSeek-V4-Flash or Qwen3.5 4B?

DeepSeek-V4-Flash vs Qwen3.5 4B

25 SHARED BENCHMARKS

Across 25 shared benchmarks, DeepSeek-V4-Flash scores higher on 24 and Qwen3.5 4B on 1. The widest gap is gdpval, where DeepSeek-V4-Flash scores 52.9 against 8.3. Qwen3.5 4B is the cheaper of the two on tracked API pricing ($0.03 against $0.44 per million input tokens).

DEEPSEEKVSALIBABA25 SHARED24–1 HEAD-TO-HEAD

BenchmarkDeepSeek-V4-FlashMARGINQwen3.5 4B

AA Agentic Index61.336.3

AA Intelligence49.920

AA-LCR79.763

AA-Omniscience-14.3-58.4

aime_202695.899

Artificial Analysis Coding Index69.122.6

critpt16.60

gdpval52.98.3

GPQA Diamond90.877.1

HLE38.69.9

hmmt_feb_202694.884.1

IFBench79.259.2

LiveCodeBench v690.955.8

MMLU-Pro86.479.1

MMMU-Pro74.865.4

OmniScience Accuracy40.415.1

OmniScience Non-Hallucination15.613.4

PinchBench91.371.3

scicode50.318.3

TauBench V3 - Banking39.46.8

Terminal-Bench 2.182.725.8

Terminal-Bench Hard38.618.2

τ²-Bench Telecom (AA run)95.692.1

τ³-Bench26.78.2

τ³-Bench Banking22.95.5

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. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.

Compare DeepSeek-V4-Flash withALL PAIRINGS →

DeepSeekDeepSeek-V4-Pro DeepSeek-V3 DeepSeek-V3.2

Across the fieldGemini 3.6 Flash Nemotron 3 Super 120B A12B GPT-5.5 Claude Opus 4.8 GPT-4o GPT-5.6 Sol Gemini 3.1 Pro GPT-5.4

Compare Qwen3.5 4B 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 fieldClaude Opus 4.7 Gemini 2.5 Pro Gemini 1.5 Pro Claude Sonnet 4 Claude Sonnet 5 Muse Spark 1.1 GLM-5.1 GLM-5.2