StableBeluga-7B: benchmark results
Stability AI Stable Beluga 7B is an Orca-style instruction fine-tune of Llama 2 7B. It uses the publisher's System/User/Assistant prompt format and remains separate from the pretrained Llama 2 checkpoint and the original LLaMA-based Stable Beluga 1 delta release. Provider: Stability AI. Access: Open.
Aggregate Rating 1324 ± 15, rank #8272 of 10916 rated models, from 69 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MMLU-by-task - High School Statistics | 48.15 | Accuracy (%) | 87.1 |
| MMLU-by-task - Conceptual Physics | 49.36 | Accuracy (%) | 86.2 |
| MMLU-by-task - Clinical Knowledge | 61.89 | Accuracy (%) | 85.3 |
| MMLU-by-task - Medical Genetics | 63 | Accuracy (%) | 84.2 |
| MMLU-by-task - College Chemistry | 43 | Accuracy (%) | 83.5 |
| MMLU-by-task - Management | 74.76 | Accuracy (%) | 82.1 |
| MMLU-by-task - Machine Learning | 40.18 | Accuracy (%) | 81.7 |
| MMLU-by-task - High School European History | 70.91 | Accuracy (%) | 81.6 |
| MMLU-by-task - Human Sexuality | 64.12 | Accuracy (%) | 79.1 |
| MMLU-by-task - TruthfulQA MC1 | 34.64 | Accuracy (%) | 78.4 |
| MMLU-by-task - TruthfulQA MC2 | 50.19 | Accuracy (%) | 76.8 |
| MMLU-by-task - Anatomy | 49.63 | Accuracy (%) | 75.9 |
Interactive version: theaggregate.ai/model?slug=stablebeluga-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.