Claude 2.1

Released2023-11-21
Price per 1MNot listed
Overall score30.2 (early, few results)
ProviderAnthropic
WeightsProprietary

Scores

DimensionScoreRankBenchmarks
Overall: Balanced 30.2 2
Overall: Coding 30.7 1
Overall: Reasoning 30.1 2
Overall: Agentic work 30.7 1
Reasoning 30.7 1
Math 29.4 1

Evidence

Reasoning 30.7 ±4.6

BenchmarkResultSourceDate
GPQA Diamond 33.0% Epoch AI 2025-01-27

Math 29.4 ±3.6

BenchmarkResultSourceDate
OTIS Mock AIME 2024-2025 1.9% Epoch AI 2025-03-07

Names matched to this model

claude-2-1

Compare Claude 2.1 with other models