Which is better, GLM-5.1 or Qwen3.6 Plus?

Across 47 shared benchmarks tracked on Vector Wire, GLM-5.1 leads on 29 and Qwen3.6 Plus leads on 16 — GLM-5.1 leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

GLM-5.1 vs Qwen3.6 Plus

Z.aivsAlibaba47 shared benchmarks29–16 head-to-head

Benchmark	GLM-5.1	Qwen3.6 Plus
AA Agentic Index	66	29
AA Intelligence	41	40.5
AA-LCR	68	72.3
AA-Omniscience	0.8	2.6
AgentWorldBench - Android	59.1	57.6
AgentWorldBench - MCP	67.6	55.3
AgentWorldBench - OS	59.1	60.3
AgentWorldBench - Overall	51.3	50.8
AgentWorldBench - Search	22.5	21.9
AgentWorldBench - SWE	52.1	59.1
AgentWorldBench - Terminal	47.3	50.6
AgentWorldBench - Web	51.5	50.8
aime_2026	95.8	95.3
arena_elo	1468	1444
arena_text_factuality	1458	1438
Artificial Analysis Coding Index	55.8	54.5
coding_arena_elo	1510	1460
critpt	4.6	2.9
Finance Agent v2	44.8	40.9
frontiermath_tier_4	12.5	8.3
gdpval	49.5	32
GPQA Diamond	86.8	90.4
HLE	52.3	28.8
HLE (with tools)	52.3	50.6
HMMT 2025	94	94.6
hmmt_feb_2026	89.4	87.8
hmmt_nov_2025	94	94.6
IFBench	76.3	75.2
imo_answer_bench	83.8	83.8
livebench	70.2	70.8
MCP Atlas	75.6	74.1
MMLU-Pro	86	88.5
nl2repo	42.7	37.9
OmniScience Accuracy	25.2	26.4
OmniScience Non-Hallucination	70.1	68
scicode	43.8	40.7
simpleqa_verified	38.1	49.1
SWE-bench Multilingual	73.3	73.8
SWE-bench Pro	58.4	56.6
TauBench V3 - Banking	13.6	20.8
Terminal-Bench 2.0	69	61.6
Terminal-Bench 2.1	63.5	61.4
Terminal-Bench Hard	43.2	43.9
tool_decathlon	40.7	39.8
toolathlon	40.7	39.8
τ²-Bench Telecom (AA run)	97.7	97.7
τ³-Bench	70.6	70.7

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.