Llama 3 15B Instruct: benchmark results
Provider: Meta. Access: Open.
Aggregate Rating 1393 ± 40, rank #6422 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 67.3 | Accuracy (%) (5-shot) | 89.2 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 53 | MC2 (%) (0-shot) | 58.1 |
| Open LLM Leaderboard v1 - GSM8K | 23.65 | Accuracy (%) (5-shot) | 43 |
| Open LLM Leaderboard v1 - WinoGrande | 74.51 | Accuracy (%) (5-shot) | 37 |
| Open LLM Leaderboard v1 - ARC Challenge | 55.97 | Normalized accuracy (%) (25-shot) | 34.6 |
| Open LLM Leaderboard v1 - HellaSwag | 74.63 | Normalized accuracy (%) (10-shot) | 25 |
Interactive version: theaggregate.ai/model?slug=llama-3-15b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.