Grok Beta

xAI

Verified against Artificial Analysis · Jul 13, 2026

Capability

percentile index

Price /M

$0 / $0

in / out

Context

—

Avg score

41.6

6 benchmarks

AA Intelligence

7.5

index

Weights

Closed

Released

Aug 13, 2024

About

Grok Beta is a proprietary model in the Grok family from xAI. Benchmarked on 6 evals, averaging 41.6.

Benchmark scores · 6

avg 41.6 — higher is better

MATHartificial-analysis · 2026-07

73.7

MMLU-Proartificial-analysis · 2026-07

70.3

GPQA Diamondartificial-analysis · 2026-07

47.1

SciCodeartificial-analysis · 2026-07

29.5

LiveCodeBenchartificial-analysis · 2026-07

24.1

Humanity's Last Examartificial-analysis · 2026-07

4.7

Details

PublisherxAI

FamilyGrok

Parameters—

License—

WeightsClosed

ReleasedAug 13, 2024

History

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$0/M