Which is better, Qwen3.6 Plus or Qwen3.7 Max?

Across 55 shared benchmarks tracked on Vector Wire, Qwen3.6 Plus leads on 4 and Qwen3.7 Max leads on 50 — Qwen3.7 Max leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

Qwen3.6 Plus vs Qwen3.7 Max

AlibabavsAlibaba55 shared benchmarks4–50 head-to-head

Benchmark	Qwen3.6 Plus	Qwen3.7 Max
AA Agentic Index	29	30.9
AA Intelligence	40.5	46.7
AA-LCR	72.3	74.7
AA-Omniscience	2.6	13.5
aime_2026	95.3	97
arena_elo	1444	1474
arena_text_factuality	1438	1479
Artificial Analysis Coding Index	54.5	66
Claw Eval (pass@3)	58.7	65.2
coding_arena_elo	1460	1517
critpt	2.9	13.4
Finance Agent v2	40.9	48.4
gdpval	32	38.5
GPQA Diamond	90.4	92.4
HLE	28.8	41.4
HLE (with tools)	50.6	53.5
hmmt_feb_2026	87.8	97.1
hmmt_nov_2025	94.6	95
IFBench	75.2	80.5
ifeval	94.3	94.3
imo_answer_bench	83.8	90
include	85.1	86.2
livebench	70.8	74.3
livebench_agentic_coding	41.4	43.6
livebench_coding	78.2	74.2
livebench_data_analysis	69.9	71.8
livebench_instruction_following	58.3	74
livebench_language	75	79.7
livebench_math	83.7	85.3
livebench_reasoning	75.8	83.3
LiveCodeBench v6	87.1	91.6
longbench_v2	62	65.3
MCP Atlas	74.1	76.4
mcpmark	48.2	60.8
mmlu_prox	84.7	87
mmlu_redux	94.5	95
MMLU-Pro	88.5	89.6
mmmlu	89.5	90.3
nl2repo	37.9	47.2
OmniScience Accuracy	26.4	31.1
OmniScience Non-Hallucination	68	74.4
scicode	40.7	53.5
simpleqa_verified	49.1	58.5
supergpqa	71.6	73.6
SWE-bench Multilingual	73.8	78.3
SWE-bench Pro	56.6	60.6
SWE-bench Verified	78.8	80.4
TauBench V3 - Banking	20.8	11.8
Terminal-Bench 2.0	61.6	69.7
Terminal-Bench 2.1	61.4	75
Terminal-Bench Hard	43.9	50.8
VITA-Bench	44.3	47.9
WideSearch	74.3	75.2
τ²-Bench Telecom (AA run)	97.7	94.7
τ³-Bench	70.7	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.