Atlas

Models

← All models

Grok 2 1212

xAI · 2024-12-12

Grok 2 1212 is xAI's December 12, 2024 update to Grok 2 with improved accuracy, instruction following, and multilingual capabilities.

Benchmark scores

BenchmarkScore
Artificial Analysis Intelligence Index8.0
BuzzBench51.8
Capability125.6
Epoch Capabilities Index (ECI)131
GPQA Diamond53.8
GPQA Diamond (Artificial Analysis avg@5)51.0
Humanity's Last Exam (Text-Only)3.8
LiveCodeBench (Artificial Analysis)26.7
MATH Level 563.5
MATH-500 (Artificial Analysis source)77.8
MMLU Pro70.9
OTIS Mock AIME 2024-202511.5
SciCode (Artificial Analysis avg@3)28.5
SimpleBench22.7
TaxEval v267.0
TaxEval v2 - Correctness54.9
TaxEval v2 - Stepwise Reasoning79.2
ZeroEval62.8
ZeroEval CRUX75.3
ZeroEval MATH Level 560.9
ZeroEval MMLU-Redux87.4
ZeroEval ZebraLogic27.7
Loading Atlas data…