Lexora-Medium-7B — benchmark results
DeepMount00's Italian-focused instruction fine-tune of a Qwen2.5 7B base, trained on his Sonnet-3.5 ITA instruction and DPO datasets. Provider: Other. Released 2024-09-24. Access: Open.
Unified ELO 1452 ± 22, rank #988 of 1776 rated models, from 24 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Pinocchio Italian - Matematica E Scienze | 68.66 | Accuracy (%) | 100 |
| Pinocchio Italian - Diritto | 64.45 | Accuracy (%) | 97.7 |
| Pinocchio Italian - Lingua Straniera | 84.69 | Accuracy (%) | 97.7 |
| Pinocchio Italian - Logica | 56.58 | Accuracy (%) | 97.7 |
| Pinocchio Italian Leaderboard | 68.98 | Average Accuracy (%) | 97.7 |
| Pinocchio Italian - Cultura | 71.44 | Accuracy (%) | 95.5 |
| Pinocchio Italian - Generale | 68.06 | Accuracy (%) | 95.5 |
| EVALITA - lexical-substitution | 40.88 | CPS | 83 |
| Open LLM Leaderboard - MuSR | 14.76 | Score | 80 |
| Open LLM Leaderboard - MMLU-Pro | 36.95 | Score | 78.3 |
| Open LLM Leaderboard - MATH Level 5 | 22.21 | Score | 75.1 |
| EVALITA - summarization-fanpage | 28.36 | CPS | 68.1 |
Interactive version: theaggregate.ai/model?slug=lexora-medium-7b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.