CruxEval-O

AI benchmark results and model performance

Type: numericalReasoning

24.28

0.51 - 55.75

0.51 - Codestral (2025-07-30)

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Mistral	Codestral (2025-07-30)	30 Jul 2025	0.51	inferred version-family alias from codestral-22b	Yes	Source
Mistral	Codestral (2024-05-29)	29 May 2024	0.51	inferred modality/version alias from codestral-22b	Yes	Source
Mistral	Codestral (2025-01-13)	13 Jan 2025	0.51	inferred modality/version alias from codestral-22b	Yes	Source
IBM	Granite 4.1 3B	29 Apr 2026	40.75	pass@1	Yes	Source
IBM	Granite 4.1 8B	29 Apr 2026	47.63	pass@1	Yes	Source
IBM	Granite 4.1 30B	29 Apr 2026	55.75	pass@1	Yes	Source