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 Sonnet 5 vs GPT-5.4

AnthropicvsOpenAI39 shared benchmarks2019 head-to-head
BenchmarkClaude Sonnet 5GPT-5.4
AA Agentic Index49.758.2
AA Intelligence55.353.1
AA-LCR7777.7
AA-Omniscience16.45.8
arena_elo14631477
arena_text_factuality14531476
Artificial Analysis Coding Index71.571.1
browsecomp84.782.7
charxiv_rq7782.8
coding_arena_elo15401457
critpt16.923.4
DeepSWE 1.15452
gdpval54.850.1
GPQA Diamond91.193
HLE57.443.7
Legal Agent Benchmark5.80.4
livebench_agentic_coding59.453.8
livebench_coding80.777.5
livebench_data_analysis71.779.3
livebench_instruction_following63.970.2
livebench_language7582.6
livebench_math92.994.2
livebench_reasoning88.788.1
MMMU-Pro77.381.2
officeqa73.368.1
officeqa_pro59.451.1
OmniScience Accuracy4050.9
OmniScience Non-Hallucination60.617.4
OSWorld-Verified81.275
scicode53.656.6
simpleqa_verified2545.3
SWE-bench Multilingual78.371.7
SWE-bench Pro63.259.1
TauBench V3 - Banking37.339.6
Terminal-Bench 2.080.475.1
Terminal-Bench 2.180.578.3
toolathlon54.354.6
usamo_202679.595.2
τ³-Bench28.272.9

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.