Phi-3 (14B): benchmark results
Provider: Microsoft. Access: Open.
Aggregate Rating 1514 ± 24, rank #2130 of 10916 rated models, from 86 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM MMLU - International Law | 93.39 | EM | 91 |
| HELM MMLU - High School European History | 88.48 | EM | 89.7 |
| HELM MMLU - Us Foreign Policy | 95 | EM | 83.3 |
| HELM Lite - GSM8K | 87.8 | Exact Match (%) | 80.9 |
| HELM MMLU - Medical Genetics | 91 | EM | 80.1 |
| HELM MMLU - Machine Learning | 69.64 | EM | 78.8 |
| HELM MMLU - Jurisprudence | 87.96 | EM | 76.9 |
| HELM MMLU - High School Mathematics | 50.37 | EM | 75.6 |
| HELM MMLU - College Mathematics | 53 | EM | 73.7 |
| HELM MMLU - High School Physics | 61.59 | EM | 73.7 |
| HELM MMLU - Elementary Mathematics | 70.9 | EM | 73.1 |
| HELM MMLU - High School Microeconomics | 90.76 | EM | 73.1 |
Interactive version: theaggregate.ai/model?slug=phi-3-14b · How It Works · Data refreshed daily, snapshot 2026-11-08.