llama-3-chinese-8B-instruct-v3: benchmark results
Provider: Meta. Access: Open.
Aggregate Rating 1428 ± 25, rank #4778 of 10916 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 67.9 | Accuracy (%) (5-shot) | 90.2 |
| Open LLM Leaderboard v1 - GSM8K | 59.21 | Accuracy (%) (5-shot) | 70.9 |
| Open Chinese LLM - C-Eval Semantic | 74.19 | Accuracy (%) | 63 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.4 | Normalized accuracy (%) (25-shot) | 62.6 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 53.57 | MC2 (%) (0-shot) | 60 |
| Open Chinese LLM - CMMLU | 54.39 | Accuracy (%) | 57.1 |
| Open Chinese LLM Leaderboard | 57.25 | Official Nonzero-Task Mean (%) | 52.1 |
| Open LLM Leaderboard v1 - WinoGrande | 76.24 | Accuracy (%) (5-shot) | 47 |
| Open LLM Leaderboard v1 - HellaSwag | 80.51 | Normalized accuracy (%) (10-shot) | 41.2 |
| Open Chinese LLM - ARC Challenge | 49.32 | Accuracy (%) | 38.5 |
| Open Chinese LLM - GSM8K | 38.67 | Accuracy (%) | 36.2 |
| Open Chinese LLM - WinoGrande | 62.04 | Accuracy (%) | 35.7 |
Interactive version: theaggregate.ai/model?slug=llama-3-chinese-8b-instruct-v3 · How It Works · Data refreshed daily, snapshot 2026-11-08.