Llama 3.1 70B Instruct: benchmark results
Meta's mid-size open Llama 3.1 instruction-tuned model (70B dense) with a 128K context and multilingual support (July 2024). Provider: Meta. Released 2024-07-23. Access: Open.
Unified ELO 1538 ± 1, rank #490 of 1392 rated models, from 825 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| European LLM Leaderboard - Accuracy | 70.57 | Average Accuracy (%) | 100 |
| HELM SeaHELM - UIT-VSFC | 73.08 | Macro F1 score | 100 |
| HREF | 48.98 | Average HREF Score (%) | 100 |
| HumanLikeness - Meaning-1 | 80.22 | Humanlike Score (%) | 100 |
| HumanLikeness - Overall | 66.91 | Overall Humanlike (%) | 100 |
| HumanLikeness - Sound-1 | 88.51 | Humanlike Score (%) | 100 |
| LegalEvalHub - LegalBench | 75.8 | Raw Metric Average (%, 162 tasks) | 100 |
| NVIDIA LLM Robustness - AGIEval Accuracy | 70.81 | Accuracy (%) | 100 |
| NVIDIA LLM Robustness - MMLU-Pro Accuracy | 64.16 | Accuracy (%) | 100 |
| Open LLM Leaderboard - IFEval | 86.69 | Score | 100 |
| RedCode | 75.54 | Score (%) | 100 |
| SCORE Robustness (Accuracy) | 67.02 | Average Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=llama-3-1-70b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-05.