OLMo 2 7B: benchmark results

Provider: Allen AI. Released 2024-11-01. Access: Open.

Unified ELO 1364 ± 1, rank #1233 of 1392 rated models, from 22 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Humanity's Last Exam5.38Accuracy (%)35.2
MIST (Selective Trust)68.1Overall Accuracy (%)22.7
Artificial Analysis Intelligence Index1Intelligence Index9.4
AA IFBench24.42Accuracy (%)6.7
AA Long Context Reasoning0Accuracy (%)6.1
AA Terminal-Bench Hard0Accuracy (%)5.5
AI for Education Pedagogy - Technology47.17Accuracy (%)5.3
AA GPQA Diamond28.79Accuracy (%)4.9
AI for Education SEND46.33Accuracy (%)4.3
AI for Education Pedagogy - Science33.33Accuracy (%)4
AI for Education Pedagogy36.48Accuracy (%)3.8
AI for Education Pedagogy - Primary36.15Accuracy (%)3.8

Interactive version: theaggregate.ai/model?slug=olmo-2-7b · How It Works · Data refreshed daily, snapshot 2026-09-05.