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 Gemini 3.6 Flash

AnthropicvsGoogle41 shared benchmarks2516 head-to-head
BenchmarkClaude Sonnet 5Gemini 3.6 Flash
AA Agentic Index49.740.5
AA Intelligence55.352
AA-LCR7779
AA-Omniscience16.422.1
Agent's Last Exam (Pass rate)33.324.2
arena_elo14631481
arena_text_factuality14531470
Artificial Analysis Coding Index71.569.2
charxiv_reasoning_with_tools88.389.4
charxiv_rq7785.2
Code Arena (Elo)15411538
coding_arena_elo15401538
critpt16.910.6
DeepSWE 1.15449
GDP (Surge AI)81.622
gdpval54.846.1
GDPval-AA v216181421
GDPVal-AA v2 (Elo)15981422
GPQA Diamond91.192.8
HLE57.440.8
HLE-Verified3151.2
livebench_agentic_coding59.443.4
livebench_coding80.777.9
livebench_data_analysis71.763
livebench_instruction_following63.975.4
livebench_language7583.9
livebench_math92.986.4
livebench_reasoning88.785.2
LVBench68.584.2
MLE-Bench66.963.9
MMMU-Pro77.383.2
mrcr_v2_8needle_128k_average81.591.8
OmniScience Accuracy4050
OmniScience Non-Hallucination60.644.4
OSWorld-Verified81.283
scicode53.652.7
simpleqa_verified2568.7
SWE-bench Pro63.258.7
TauBench V3 - Banking37.329.9
Terminal-Bench 2.180.578
Terminal-bench 3.014.65.4

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.