Which is better, Gemini 3.1 Pro or Qwen3.7 Max?

Across 50 shared benchmarks tracked on Vector Wire, Gemini 3.1 Pro leads on 31 and Qwen3.7 Max leads on 19 — 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 Qwen3.7 Max

GooglevsAlibaba50 shared benchmarks31–19 head-to-head

Benchmark	Gemini 3.1 Pro	Qwen3.7 Max
AA Agentic Index	23	30.9
AA Intelligence	47.7	46.7
AA-LCR	79	74.7
AA-Omniscience	32.9	13.5
aime_2026	98.3	97
arena_elo	1486	1474
arena_text_factuality	1471	1479
Artificial Analysis Coding Index	68.8	66
coding_arena_elo	1447	1517
critpt	17.7	13.4
DeepSWE	10	18
DeepSWE 1.1	12	21.6
Finance Agent v2	43	48.4
gdpval	23.3	38.5
GPQA Diamond	94.3	92.4
HLE	51.4	41.4
HLE (with tools)	51.6	53.5
hmmt_feb_2026	94.7	97.1
hmmt_nov_2025	94.8	95
IFBench	77.1	80.5
imo_answer_bench	91	90
itbenchSre	30.3	42.5
livebench	79.9	74.3
livebench_agentic_coding	44.1	43.6
livebench_coding	76.5	74.2
livebench_data_analysis	78.5	71.8
livebench_instruction_following	79.1	74
livebench_language	85.4	79.7
livebench_math	91	85.3
livebench_reasoning	84	83.3
LiveCodeBench v6	91.7	91.6
MCP Atlas	78.2	76.4
mcpmark	55.9	60.8
MMLU-Pro	91	89.6
mmmlu	92.6	90.3
nl2repo	33.4	47.2
OmniScience Accuracy	55.3	31.1
OmniScience Non-Hallucination	50.1	74.4
scicode	59	53.5
simplebench	79.6	70.4
simpleqa_verified	77.3	58.5
SWE-bench Multilingual	76.9	78.3
SWE-bench Pro	54.2	60.6
SWE-bench Verified	80.6	80.4
TauBench V3 - Banking	21.4	11.8
Terminal-Bench 2.0	68.5	69.7
Terminal-Bench 2.1	74	75
Terminal-Bench Hard	68.5	50.8
τ²-Bench Telecom (AA run)	99.3	94.7
τ³-Bench	67.1	10.9

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.