Mistral-7B-v0.3 — benchmark results

Provider: Mistral. Released 2024-05-22. Access: Open.

Unified ELO 1369 ± 15, rank #1346 of 1776 rated models, from 97 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
EuroEval Finnish NLU - Tydiqa FI69.63Reading comprehension Score (%)83.6
EuroEval Portuguese NLU - MultiWikiQA PT74.69Reading comprehension Score (%)83.4
EuroEval Spanish NLU - MLQA ES64.36Reading comprehension Score (%)79.1
LIBRA - ru2WikiMultihopQA41Dataset Total Score (%)78.6
LIBRA - ruQasper5.75Dataset Total Score (%)78.6
EuroEval Dutch NLU - DBRD90.5Sentiment classification Score (%)76.2
LA Leaderboard - PIQA Catalan65.13Accuracy (%)75.7
EuroEval Portuguese NLU - SST-2 PT80.72Sentiment classification Score (%)70.7
LIBRA - LongContextMultiQ5.25Dataset Total Score (%)64.3
LIBRA - ruBABILongQA315.67Dataset Total Score (%)64.3
LIBRA - ruGSM1009Dataset Total Score (%)64.3
EuroEval Italian NLU - SQuAD IT69.12Reading comprehension Score (%)63

Interactive version: theaggregate.ai/model?slug=mistral-7b-v0-3 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.