Which is better, Gemini 3 Pro or Qwen3.5 27B?

Across 53 shared benchmarks tracked on Vector Wire, Gemini 3 Pro leads on 38 and Qwen3.5 27B leads on 15 — Gemini 3 Pro leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

Gemini 3 Pro vs Qwen3.5 27B

GooglevsAlibaba53 shared benchmarks38–15 head-to-head

Benchmark	Gemini 3 Pro	Qwen3.5 27B
AA Agentic Index	52	54.6
AA Intelligence	40.6	34.6
AA-LCR	73	72.3
AA-Omniscience	15.3	-44
aime_2026	91.7	92.6
arena_vision	1290	1221
Artificial Analysis Coding Index	46.5	34.9
browsecomp	59.2	61
browsecomp_zh	66.8	62.1
charxiv_rq	81.4	79.5
coding_arena_elo	1438	1357
critpt	9.1	0.9
gdpval	34.2	33
GPQA Diamond	91.9	85.8
HLE	45.8	48.5
hmmt_feb_2025	97.5	92
hmmt_feb_2026	86.4	84.3
hmmt_nov_2025	93.3	89.8
IFBench	70.4	76.5
imo_answer_bench	83.3	79.9
LiveCodeBench v6	90.7	80.7
longbench_v2	68.2	60.6
LVBench	73.5	73.6
mathvision	86.1	86
MathVista	89.8	87.8
MCP Atlas	70.3	68.4
MLVU	80.7	85.9
MMLU-Pro	90.1	86.1
mmmlu	91.8	85.9
MMMU-Pro	81	75
MMVU	78.9	73.3
multichallenge	65.7	60.8
OCRBench	90.3	89.4
OmniDocBench 1.5	88.5	88.9
OmniScience Accuracy	55.8	20.7
OmniScience Non-Hallucination	10	24.9
scicode	56.1	39.5
screenspot_pro_no_tools	72.7	70.3
Seal-0	45.5	47.2
SimpleVQA	69.7	56
SWE-bench Multilingual	68.7	69.3
SWE-bench Pro	43.3	51.2
SWE-bench Verified	78	75
Terminal-Bench 2.0	54.2	41.6
Terminal-Bench Hard	56.9	32.6
tool_decathlon	36.4	31.5
vectara_answer_rate	99.4	99.8
vectara_avg_summary_length	101.9	94.4
vectara_factual_consistency	86.4	87.9
vectara_hallucination_rate ↓	13.6	12.1
VideoMMMU	87.6	82.3
ZeroBench	8	0.1
τ²-Bench Telecom (AA run)	98	93.9

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.