Which is better, Claude Opus 4.6 or Qwen3.5 27B?

Across 48 shared benchmarks tracked on Vector Wire, Claude Opus 4.6 leads on 35 and Qwen3.5 27B leads on 12 — Claude Opus 4.6 leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

Claude Opus 4.6 vs Qwen3.5 27B

AnthropicvsAlibaba48 shared benchmarks35–12 head-to-head

Benchmark	Claude Opus 4.6	Qwen3.5 27B
AA Agentic Index	67.6	54.6
AA Intelligence	44.9	34.6
AA-LCR	74.3	72.3
AA-Omniscience	13.7	-44
aime_2026	96.7	92.6
arena_vision	1300	1221
Artificial Analysis Coding Index	48.1	34.9
baby_vision	14.8	44.6
browsecomp	86.8	61
charxiv_rq	69.1	79.5
coding_arena_elo	1545	1357
critpt	12.6	0.9
ERQA	40.8	60.5
gdpval	55.9	33
GPQA Diamond	91.3	85.8
HLE	53.1	48.5
hmmt_feb_2026	96.2	84.3
hmmt_nov_2025	96.3	89.8
IFBench	62.5	76.5
imo_answer_bench	75.3	79.9
LiveCodeBench v6	88.8	80.7
mathvision	71.2	86
MCP Atlas	76.8	68.4
mcpmark	56.7	36.3
MMLU-Pro	89.1	86.1
mmmlu	91.1	85.9
MMMU-Pro	77.3	75
multichallenge	56	60.8
nl2repo	49.8	27.3
OmniDocBench 1.5	86.6	88.9
OmniScience Accuracy	47	20.7
OmniScience Non-Hallucination	37.2	24.9
OSWorld-Verified	72.7	56.2
RealWorldQA	73.9	83.7
scicode	52	39.5
screenspot_pro_no_tools	57.7	70.3
SWE-bench Multilingual	77.8	69.3
SWE-bench Pro	57.3	51.2
SWE-bench Verified	80.8	75
Terminal-Bench 2.0	65.4	41.6
Terminal-Bench Hard	65.4	32.6
tool_decathlon	47.2	31.5
vectara_answer_rate	99.8	99.8
vectara_avg_summary_length	137.6	94.4
vectara_factual_consistency	87.8	87.9
vectara_hallucination_rate ↓	12.2	12.1
τ²-Bench Telecom (AA run)	99.3	93.9
τ³-Bench	72.4	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.