ministral-3B-2410: benchmark results

Provider: Mistral. Released 2024-10-16. Access: Open.

Unified ELO 1458 ± 1, rank #908 of 1392 rated models, from 12 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Vectara Hallucination Leaderboard92.7Factual Consistency Rate (%)69.2
Icelandic LLM - WinoGrande-IS54.69Score (%)20.4
Icelandic LLM - Inflection18.04Score (%)18.3
Icelandic LLM - GED51Score (%)16.7
Icelandic LLM - Belebele-IS40.89Score (%)16.1
Icelandic LLM Leaderboard - Average32.57Average Score (%)16.1
Icelandic LLM - ARC-Challenge-IS28.5Score (%)15.1
Wolfram LLM Benchmarking Project17.9Correct Functionality (%)14.4
Epoch AI - Dtbench51.73Score13.1
MATH Level 514.44Accuracy (%)12
Icelandic LLM - WikiQA-IS2.32Score (%)9.7
Conceptual Reasoning Index - Argument Evaluation (LMCA)5.48Chance-Corrected Score (0-100)6.1

Interactive version: theaggregate.ai/model?slug=ministral-3b-2410 · How It Works · Data refreshed daily, snapshot 2026-09-05.