Llama 2 13B: benchmark results
Meta's Llama 2 13B, the base pretrained checkpoint released under the Llama 2 Community License in July 2023; the linked model card provides no further details here. Provider: Meta. Released 2023-07-18. Access: Open.
Aggregate Rating 1322 ± 10, rank #1436 of 1656 rated models, from 191 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - IMDB | 96.2 | Exact Match (%) | 98.5 |
| HELM Classic - Entity Data Imputation | 84.36 | Exact Match (%) | 97.7 |
| ToolBench - WebShop Short | 31.67 | Task Score | 97.7 |
| HELM Classic - NarrativeQA | 74.4 | F1 (%) | 93.8 |
| HELM Classic - LSAT | 23.48 | Exact Match (%) | 92.6 |
| HELM Classic - WikiFact | 36.47 | Exact Match (%) | 90.9 |
| SAD - Predict Words | 48.67 | Self-prediction score (%, higher is better) | 90.9 |
| SAD - Predict Words (Situating Prompt) | 47.44 | Self-prediction score (%, higher is better) | 90.9 |
| ToolBench - WebShop Long | 0.6 | Task Score | 90.7 |
| SAD - Anti-Imitation | 19.89 | Score (%, higher is better) | 90 |
| SAD - Anti-Imitation (Situating Prompt) | 20.59 | Score (%, higher is better) | 90 |
| HELM | 82.3 | Mean win rate (imported) | 89.2 |
Interactive version: theaggregate.ai/model?slug=llama-2-13b · How It Works · Data refreshed daily, snapshot 2026-10-20.