miqu-1-70B-sf — benchmark results

Community dequantization of 'miqu-1-70b', a leaked early Mistral Medium prototype retrained from Llama 2 70B, whose origin Mistral's CEO confirmed (February 2024). Provider: Other. Released 2024-02-01. Access: Open.

Unified ELO 1503 ± 7, rank #787 of 1776 rated models, from 106 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
EuroEval Spanish NLU - Sentiment Headlines ES51.27Sentiment classification Score (%)94.5
EuroEval Swedish NLU - Swerec79.32Sentiment classification Score (%)92.6
Open LLM Leaderboard - MuSR17.21Score88
Open LLM Leaderboard - GPQA13.42Score87.9
EuroEval English NLU - SST-568.54Sentiment classification Score (%)84.6
EuroEval Finnish NLU - Scandisent FI91.99Sentiment classification Score (%)84.4
Open LLM Leaderboard - BBH43.81Score83.2
EuroEval Italian NLU - ScaLA IT32.89Linguistic acceptability Score (%)80.8
EuroEval Portuguese NLU - ScaLA PT24.01Linguistic acceptability Score (%)78
Open LLM Leaderboard - MMLU-Pro35.87Score76.8
EuroEval Italian NLU - Sentipolc1660.69Sentiment classification Score (%)76.4
EuroEval Portuguese56.68Average Score (%)74.9

Interactive version: theaggregate.ai/model?slug=miqu-1-70b-sf · How the rankings work · Data refreshed daily, snapshot 2026-07-22.