DBRX Instruct — benchmark results

Provider: Databricks. Released 2024-03-27. Access: Open.

Unified ELO 1248 ± 38, rank #1621 of 1776 rated models, from 20 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Humanity's Last Exam6.59Accuracy (%)48.4
HELM Lite32.55Mean win rate (self-reported)26
HELM (Stanford)28.87Mean Win Rate (%)21.1
HELM Safety BBQ79.2BBQ accuracy (%)18.6
HELM NaturalQuestions (Closed)28.39F1 (%)15.6
HELM WMT 201413.12BLEU-4 (%)12.2
AA SciCode11.81Accuracy (%)12
AA GPQA Diamond33.13Accuracy (%)11.3
HELM NarrativeQA48.84F1 (%)7.8
AA MMLU-Pro39.71Accuracy (%)7.3
Artificial Analysis Intelligence Index2.96Intelligence Index7.1
AA LiveCodeBench9.31Pass@1 (%)6.4

Interactive version: theaggregate.ai/model?slug=dbrx-instruct · How the rankings work · Data refreshed daily, snapshot 2026-07-22.