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 Fable 5 vs Claude Sonnet 4.6

AnthropicvsAnthropic54 shared benchmarks504 head-to-head
BenchmarkClaude Fable 5Claude Sonnet 4.6
AA Agentic Index56.661.6
AA Intelligence62.148.4
AA-LCR76.774
AA-Omniscience43.312.2
arc_agi_190.586.5
ARC-AGI-276.860.4
arena_elo15081472
arena_text_factuality14861460
arena_vision13071278
Artificial Analysis Coding Index76.563
Blueprint-Bench 238.66.7
browsecomp8876.2
charxiv_rq89.472.4
coding_arena_elo16261523
critpt28.63.1
DeepSWE 1.17030
Finance Agent v256.351
GDP (Surge AI)29.878.6
gdpval61.954.8
GDPval-AA (Elo)19321676
GDPval-AA v217601381
GPQA Diamond92.689.9
healthbench_professional6644.2
HLE64.549
HLE (with tools)64.546.8
IFBench63.556.6
Legal Agent Benchmark13.35.4
Legal Agent Benchmark (Harvey's Held-Out Set)13.35.4
livebench78.375.5
livebench_agentic_coding62.242.6
livebench_coding8679.3
livebench_data_analysis80.578
livebench_instruction_following75.863.2
livebench_language90.776.1
livebench_math9687
livebench_reasoning89.784.8
MCP Atlas84.769.5
MMMU-Pro84.275.6
officeqa_pro69.953.4
OmniScience Accuracy65.340.9
OmniScience Non-Hallucination36.451.6
OSWorld-Verified8578.5
scicode60.247
simpleqa_verified68.329
SWE-bench Pro8058.1
SWE-bench Verified9579.6
TauBench V3 - Banking38.134.4
Terminal-Bench 2.18871.2
Terminal-Bench Hard62.959.1
toolathlon61.749.4
Toolathlon Avg turns19.816.5
Toolathlon Pass@∞55.638
τ²-Bench Telecom (AA run)98.597.9
τ³-Bench26.830.5

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.