Llama 3 8B Instruct V0.8 [float16] [rev: A835be513190]: benchmark results
Provider: MaziyarPanahi. Access: Open.
Aggregate Rating 1462 ± 35, rank #3224 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 68.3 | Accuracy (%) (5-shot) | 91.1 |
| Open LLM Leaderboard v1 - ARC Challenge | 71.93 | Normalized accuracy (%) (25-shot) | 90.5 |
| Open LLM Leaderboard v1 - HellaSwag | 87.75 | Normalized accuracy (%) (10-shot) | 87.7 |
| Open LLM Leaderboard v1 - GSM8K | 68.16 | Accuracy (%) (5-shot) | 85.2 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 63.94 | MC2 (%) (0-shot) | 81.4 |
| Open LLM Leaderboard v1 - WinoGrande | 78.93 | Accuracy (%) (5-shot) | 66.3 |
Interactive version: theaggregate.ai/model?slug=llama-3-8b-instruct-v0-8-float16-rev-a835be513190 · How It Works · Data refreshed daily, snapshot 2026-11-08.