Which is better, Qwen3.6 35B A3B or Qwen3.7 Max?

Across 36 shared benchmarks tracked on Vector Wire, Qwen3.6 35B A3B leads on 2 and Qwen3.7 Max leads on 34 — 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 35B A3B vs Qwen3.7 Max

AlibabavsAlibaba36 shared benchmarks2–34 head-to-head

Benchmark	Qwen3.6 35B A3B	Qwen3.7 Max
AA Agentic Index	21.6	30.9
AA Intelligence	32.1	46.7
AA-LCR	66.7	74.7
AA-Omniscience	-22.2	13.5
aime_2026	92.7	97
Artificial Analysis Coding Index	41.9	66
Claw Eval (pass@3)	50	65.2
critpt	0.3	13.4
gdpval	27.8	38.5
GPQA Diamond	86	92.4
HLE	22.2	41.4
hmmt_feb_2026	83.6	97.1
hmmt_nov_2025	89.1	95
IFBench	64.4	80.5
imo_answer_bench	78.9	90
LiveCodeBench v6	80.4	91.6
MCP Atlas	62.8	76.4
mcpmark	37	60.8
mmlu_redux	93.3	95
MMLU-Pro	85.6	89.6
nl2repo	29.4	47.2
OmniScience Accuracy	18.8	31.1
OmniScience Non-Hallucination	49.5	74.4
scicode	35.8	53.5
supergpqa	64.7	73.6
SWE-bench Multilingual	67.2	78.3
SWE-bench Pro	49.5	60.6
SWE-bench Verified	73.4	80.4
TauBench V3 - Banking	9.3	11.8
Terminal-Bench 2.0	51.5	69.7
Terminal-Bench 2.1	44.9	75
Terminal-Bench Hard	34.8	50.8
VITA-Bench	35.6	47.9
WideSearch	60.1	75.2
τ²-Bench Telecom (AA run)	95.3	94.7
τ³-Bench	67.2	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.