VECTOR WIREAI INTELLIGENCE
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models Deals Regulatory Sources

Claude Opus 4.5 vs Claude Sonnet 5

AnthropicvsAnthropic33 shared benchmarks825 head-to-head
BenchmarkClaude Opus 4.5Claude Sonnet 5
AA Agentic Index59.649.7
AA Intelligence41.955.3
AA-LCR7677
AA-Omniscience1416.4
arena_elo14731463
arena_text_factuality14581453
Artificial Analysis Coding Index47.871.5
browsecomp3784.7
charxiv_rq68.577
coding_arena_elo14951540
critpt4.616.9
gdpval47.354.8
GPQA Diamond8791.1
HLE30.857.4
livebench_agentic_coding39.759.4
livebench_coding79.780.7
livebench_data_analysis74.471.7
livebench_instruction_following62.563.9
livebench_language81.375
livebench_math90.492.9
livebench_reasoning80.188.7
MMMU-Pro7477.3
OmniScience Accuracy46.640
OmniScience Non-Hallucination3960.6
OSWorld-Verified66.381.2
scicode5053.6
simplebench6260.6
simpleqa_verified41.825
SWE-bench Multilingual77.578.3
SWE-bench Pro57.163.2
SWE-bench Verified81.585.2
Terminal-Bench 2.059.380.4
toolathlon43.554.3

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.