Refresh Models —Deals —Regulatory —Sources —

GPT-5.4 vs Qwen3.5 2B

20 SHARED BENCHMARKS

Across 20 shared benchmarks, GPT-5.4 scores higher on 19 and Qwen3.5 2B on 1. The widest gap is HLE, where GPT-5.4 scores 43.7 against 5.

OPENAIVSALIBABA20 SHARED19–1 HEAD-TO-HEAD

BenchmarkGPT-5.4MARGINQwen3.5 2B

AA Agentic Index58.20.8

AA Intelligence396.9

AA-LCR8225.6

AA-Omniscience5.8-60.2

Artificial Analysis Coding Index71.12.9

GPQA Diamond9351.6

HLE43.75

IFBench73.941.3

MMLU-Pro87.566.5

OmniScience Accuracy50.98.1

OmniScience Non-Hallucination17.425.8

scicode56.67.2

TauBench V3 - Banking39.62.5

Terminal-Bench 2.178.33

Terminal-Bench Hard57.63.8

τ²-Bench Telecom (AA run)87.181.6

τ³-Bench72.96.2

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 GPT-5.4 withALL PAIRINGS →

OpenAIGPT-5.5 GPT-4o GPT-5.6 Sol GPT-5.2 GPT-5 o3

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 fieldGPT-5.2 Claude Opus 4.6 Claude Sonnet 4.5 Claude 3.5 Sonnet Gemini 3 Pro Claude Fable 5 GPT-5 Claude Opus 4.5