Which is better, Claude Opus 4.7 or Qwen3.6 Plus?

Across 45 shared benchmarks tracked on Vector Wire, Claude Opus 4.7 leads on 41 and Qwen3.6 Plus leads on 4 — Claude Opus 4.7 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 4.7 vs Qwen3.6 Plus

AnthropicvsAlibaba45 shared benchmarks41–4 head-to-head

Benchmark	Claude Opus 4.7	Qwen3.6 Plus
AA Agentic Index	64.6	29
AA Intelligence	55	40.5
AA-LCR	75.3	72.3
AA-Omniscience	27.3	2.6
aime_2026	95.8	95.3
arena_elo	1494	1444
arena_text_factuality	1474	1438
Artificial Analysis Coding Index	73.6	54.5
coding_arena_elo	1557	1460
critpt	12	2.9
Finance Agent v2	51.5	40.9
frontiermath_tier_4	22.9	8.3
gdpval	58.6	32
GPQA Diamond	94.2	90.4
HLE	54.7	28.8
HLE (with tools)	54.7	50.6
hmmt_feb_2026	93.9	87.8
IFBench	58.6	75.2
livebench	76.9	70.8
livebench_agentic_coding	50.7	41.4
livebench_coding	82.1	78.2
livebench_data_analysis	78.3	69.9
livebench_instruction_following	66.7	58.3
livebench_language	77.9	75
livebench_math	92.8	83.7
livebench_reasoning	87.2	75.8
MCP Atlas	79.1	74.1
mmmlu	91.5	89.5
MMMU-Pro	78.8	78.8
OmniScience Accuracy	48.9	26.4
OmniScience Non-Hallucination	57.7	68
OSWorld-Verified	82.8	62.5
scicode	54.5	40.7
screenspot_pro_no_tools	79.5	68.2
simpleqa_verified	50.6	49.1
SWE-bench Multilingual	80.5	73.8
SWE-bench Pro	64.3	56.6
SWE-bench Verified	87.6	78.8
TauBench V3 - Banking	34.6	20.8
Terminal-Bench 2.0	69.4	61.6
Terminal-Bench 2.1	83.1	61.4
Terminal-Bench Hard	54.5	43.9
toolathlon	59.3	39.8
τ²-Bench Telecom (AA run)	88.6	97.7
τ³-Bench	28.9	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.