Which is better, Claude Opus 4.8 or Qwen3.6 Plus?

Across 61 shared benchmarks tracked on Vector Wire, Claude Opus 4.8 leads on 51 and Qwen3.6 Plus leads on 10 — Claude Opus 4.8 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 4.8 vs Qwen3.6 Plus

AnthropicvsAlibaba61 shared benchmarks51–10 head-to-head

Benchmark	Claude Opus 4.8	Qwen3.6 Plus
AA Agentic Index	49.4	29
AA Intelligence	57.3	40.5
AA-LCR	73	72.3
AA-Omniscience	28.8	2.6
AgentWorldBench - Android	61.5	57.6
AgentWorldBench - MCP	54.9	55.3
AgentWorldBench - OS	66.6	60.3
AgentWorldBench - Overall	56.6	50.8
AgentWorldBench - Search	35.1	21.9
AgentWorldBench - SWE	64.1	59.1
AgentWorldBench - Terminal	59.2	50.6
AgentWorldBench - Web	54.7	50.8
aime_2026	100	95.3
arena_elo	1482	1444
arena_text_factuality	1464	1438
Artificial Analysis Coding Index	74.3	54.5
coding_arena_elo	1539	1460
critpt	20.9	2.9
Finance Agent v2	53.9	40.9
frontiermath_tier_4	31.3	8.3
gdpval	54.2	32
GPQA Diamond	93.6	90.4
HLE	57.9	28.8
HLE (with tools)	57.9	50.6
hmmt_feb_2026	96.7	87.8
hmmt_nov_2025	96.5	94.6
IFBench	62.2	75.2
imo_answer_bench	83.5	83.8
include	87.6	85.1
livebench	77.2	70.8
livebench_agentic_coding	50.5	41.4
livebench_coding	81.8	78.2
livebench_data_analysis	66	69.9
livebench_instruction_following	72	58.3
livebench_language	79.7	75
livebench_math	94.3	83.7
livebench_reasoning	89.2	75.8
longbench_v2	69.1	62
mathvision	86.7	88
MCP Atlas	83.6	74.1
MMMU-Pro	78.9	78.8
nl2repo	69.7	37.9
OmniScience Accuracy	48.8	26.4
OmniScience Non-Hallucination	60.7	68
OSWorld-Verified	83.4	62.5
scicode	53.5	40.7
screenspot_pro_no_tools	87.9	68.2
simpleqa_verified	39.5	49.1
SWE-bench Multilingual	84.4	73.8
SWE-bench Pro	69.2	56.6
SWE-bench Verified	88.6	78.8
TauBench V3 - Banking	34.2	20.8
Terminal-Bench 2.0	74.6	61.6
Terminal-Bench 2.1	85	61.4
Terminal-Bench Hard	58.3	43.9
tool_decathlon	59.9	39.8
toolathlon	59.9	39.8
Video-MME	86	84.2
WideSearch	72.9	74.3
τ²-Bench Telecom (AA run)	94.4	97.7
τ³-Bench	27.6	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.