Claude Sonnet 4.5 vs Grok 4

Verdict

Evenly split 6–6 across 12 shared benchmarks.

Claude Sonnet 4.5 6 · Grok 4 6 · higher isn't always better — check the caveats below.

Per-benchmark head-to-head

Reported scores — protocols may differ. How we compare

Sort

Claude Sonnet 4.5Grok 4

63.67
BARC-AGI
66.67
13.61
AARC-AGI-2
15.97
5.3
AChess Puzzles
24.24
60
BCybench
43
52.6
BDeepResearch Bench
47.3
42.5
AGDPval
21.1
76.43
BGPQA diamond
82.67
67.38
BMETR Time Horizons
66.58
77.76
BOTIS Mock AIME 2024-2025
83.98
45.16
ASimpleBench
52.6
46.52
BTerminal Bench
27.2
47.71
AWeirdML
45.73

Capabilities

Reasoning

Reasoning: Grok 4 leads 3–1 across 4 shared reasoning benchmarks (largest gap: SimpleBench, 52.6 vs 45.16 — single evaluator).

2 reproduced6 unverified

A higher score does not always mean a better model. Each number is task performance under a disclosed harness. Of the 24 measurements across 12 shared benchmarks, 24 single-evaluator or undisclosed-protocol reports; 21% are independently reproduced. Claude Sonnet 4.5 pricing is cross-check disputed — weigh cost claims with the same caution as benchmark wins. Check any benchmark's integrity grade before reading a win as decisive.

Frequently asked questions

Evenly split 6–6 across 12 shared benchmarks. Claude Sonnet 4.5 leads 6 benchmarks and Grok 4 leads 6. Higher isn't always better — see the integrity caveats on each benchmark.

Claude Sonnet 4.5 and Grok 4 have 12 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.

Grok 4 leads reasoning 3–1.

Follow the record

These models

Same tier

Frequently asked questions

Evenly split 6–6 across 12 shared benchmarks. Claude Sonnet 4.5 leads 6 benchmarks and Grok 4 leads 6. Higher isn't always better — see the integrity caveats on each benchmark.

Claude Sonnet 4.5 and Grok 4 have 12 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.

Grok 4 leads reasoning 3–1.