Which is better, GPT-5 or Qwen3.5 35B A3B?

Across 46 shared benchmarks tracked on Vector Wire, GPT-5 leads on 30 and Qwen3.5 35B A3B leads on 15 — GPT-5 leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

GPT-5 vs Qwen3.5 35B A3B

OpenAIvsAlibaba46 shared benchmarks30–15 head-to-head

Benchmark	GPT-5	Qwen3.5 35B A3B
AA Agentic Index	49.7	44.1
AA Intelligence	35.3	29.9
AA-LCR	76.3	68.3
AA-Omniscience	-8.7	-48.1
Artificial Analysis Coding Index	38.9	37
browsecomp	54.9	61
browsecomp_zh	65	69.5
coding_arena_elo	1419	1250
critpt	5.7	0.9
ERQA	65.7	64.8
gdpval	32.5	20.3
GPQA (unspecified)	85.7	84.2
GPQA Diamond	87.3	84.5
GSM8K	-1.8	90.1
HLE	28.5	47.4
HMMT 2025	93.3	89.2
hmmt_feb_2025	88.3	89
hmmt_nov_2025	89.2	89.2
humaneval	93.4	66.5
IFBench	73.1	72.5
imo_answer_bench	76	76.8
LiveCodeBench v6	87	74.6
mmlu_redux	95.3	93.3
MMLU-Pro	87.5	85.3
MMMU	84.2	81.4
MMMU-Pro	78.4	75.1
multichallenge	71.1	60
ocrbench_v2	55.5	65.3
OmniScience Accuracy	40.3	20.1
OmniScience Non-Hallucination	21	14.6
scicode	43	37.7
Seal-0	51.4	41.4
SWE-bench Multilingual	55.3	60.3
SWE-bench Pro	41.8	44.6
SWE-bench Verified	74.9	70
TauBench V3 - Banking	22.1	4.9
Terminal-Bench 2.0	35.2	40.5
Terminal-Bench 2.1	35.2	40.8
Terminal-Bench Hard	37.9	26.5
vectara_answer_rate	99.9	99.8
vectara_avg_summary_length	162.7	94.9
vectara_factual_consistency	85.3	89.5
vectara_hallucination_rate ↓	14.7	10.5
VideoMMMU	84.6	80.4
τ²-Bench Telecom (AA run)	86.5	89.2
τ³-Bench	19.6	68.9

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.