Which is better, Gemini 3.1 Pro or Qwen3.5 27B?

Across 45 shared benchmarks tracked on Vector Wire, Gemini 3.1 Pro leads on 40 and Qwen3.5 27B leads on 5 — Gemini 3.1 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 3.1 Pro vs Qwen3.5 27B

GooglevsAlibaba45 shared benchmarks40–5 head-to-head

Benchmark	Gemini 3.1 Pro	Qwen3.5 27B
AA Agentic Index	23	54.6
AA Intelligence	47.7	34.6
AA-LCR	79	72.3
AA-Omniscience	32.9	-44
aime_2026	98.3	92.6
arena_vision	1280	1221
Artificial Analysis Coding Index	68.8	34.9
baby_vision	51.6	44.6
browsecomp	85.9	61
charxiv_rq	89.9	79.5
coding_arena_elo	1447	1357
critpt	17.7	0.9
gdpval	23.3	33
GPQA Diamond	94.3	85.8
HLE	51.4	48.5
hmmt_feb_2026	94.7	84.3
hmmt_nov_2025	94.8	89.8
IFBench	77.1	76.5
imo_answer_bench	91	79.9
itbenchSre	30.3	35.5
LiveCodeBench v6	91.7	80.7
mathvision	89.8	86
MCP Atlas	78.2	68.4
mcpmark	55.9	36.3
MMLU-Pro	91	86.1
mmmlu	92.6	85.9
MMMU-Pro	83	75
multichallenge	71.4	60.8
nl2repo	33.4	27.3
OmniScience Accuracy	55.3	20.7
OmniScience Non-Hallucination	50.1	24.9
OSWorld-Verified	76.2	56.2
scicode	59	39.5
SWE-bench Multilingual	76.9	69.3
SWE-bench Pro	54.2	51.2
SWE-bench Verified	80.6	75
Terminal-Bench 2.0	68.5	41.6
Terminal-Bench Hard	68.5	32.6
tool_decathlon	48.8	31.5
vectara_answer_rate	99.4	99.8
vectara_avg_summary_length	107.7	94.4
vectara_factual_consistency	89.6	87.9
vectara_hallucination_rate ↓	10.4	12.1
τ²-Bench Telecom (AA run)	99.3	93.9
τ³-Bench	67.1	68.4

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.