Llama 3 8B IT: benchmark results
Provider: Meta. Released 2024-04-18.
Aggregate Rating 1404 ± 42, rank #5888 of 10916 rated models, from 8 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MT-JailBench | 71.07 | CrescendoX ASR (self-reported) | 65 |
| Pinocchio Italian - Cultura | 57.21 | Accuracy (%) | 56.8 |
| Pinocchio Italian - Diritto | 51.01 | Accuracy (%) | 56.8 |
| Pinocchio Italian - Generale | 54.12 | Accuracy (%) | 56.8 |
| Pinocchio Italian - Matematica E Scienze | 47.16 | Accuracy (%) | 56.8 |
| Pinocchio Italian Leaderboard | 51.81 | Average Accuracy (%) | 54.5 |
| Pinocchio Italian - Logica | 40.05 | Accuracy (%) | 47.7 |
| Pinocchio Italian - Lingua Straniera | 61.32 | Accuracy (%) | 36.4 |
Interactive version: theaggregate.ai/model?slug=llama-3-8b-it · How It Works · Data refreshed daily, snapshot 2026-11-08.