vicuna-13B-v1.3: benchmark results
LMSYS's 13B chat assistant, fine-tuned from LLaMA on about 125,000 ShareGPT conversations under a non-commercial research license. Provider: LMSYS. Released 2023-06-18. Access: Open.
Aggregate Rating 1316 ± 11, rank #1447 of 1656 rated models, from 129 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Classic - NaturalQuestions Open Book | 68.65 | F1 (%) | 90.8 |
| MMLU-by-task - Human Sexuality | 69.47 | Accuracy (%) | 89.5 |
| HELM Classic - CivilComments | 64.46 | Exact Match (%) | 89.4 |
| HELM Classic - TruthfulQA | 38.53 | Exact Match (%) | 87.9 |
| MMLU-by-task - TruthfulQA MC1 | 36.47 | Accuracy (%) | 87.8 |
| MMLU-by-task - Machine Learning | 43.75 | Accuracy (%) | 86.9 |
| MMLU-by-task - College Computer Science | 49 | Accuracy (%) | 86.2 |
| MMLU-by-task - TruthfulQA MC2 | 52.14 | Accuracy (%) | 85.9 |
| HELM Classic - Synthetic Reasoning Natural | 29.61 | F1 (%) | 85.3 |
| HELM Classic - Entity Matching | 87.91 | Exact Match (%) | 84.8 |
| MMLU-by-task - World Religions | 78.36 | Accuracy (%) | 83.1 |
| MMLU-by-task - Marketing | 82.48 | Accuracy (%) | 82.6 |
Interactive version: theaggregate.ai/model?slug=vicuna-13b-v1-3 · How It Works · Data refreshed daily, snapshot 2026-10-20.