Jamba 1.5 Mini: benchmark results
AI21's hybrid Mamba-Transformer MoE (52B total, 12B active) with a 256K context. Provider: AI21 Labs. Released 2024-08-22. Access: API.
Aggregate Rating 1452 ± 19, rank #983 of 1656 rated models, from 143 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RULER | 93.9 | Avg Accuracy (%) | 88.6 |
| HELMET (128K) | 46.9 | Average Score | 87.1 |
| HELM MMLU - Jurisprudence | 87.96 | EM | 76.9 |
| HELM MMLU - Virology | 57.83 | EM | 74.4 |
| NoLiMa | 92.4 | Base Score (%) | 66.7 |
| HELM MMLU - Miscellaneous | 90.17 | EM | 59.6 |
| HELM MMLU - International Law | 89.26 | EM | 57.7 |
| HELM NarrativeQA | 74.62 | F1 (%) | 57.2 |
| HELM NaturalQuestions (Closed) | 38.79 | F1 (%) | 54.4 |
| HELM Lite - OpenbookQA | 89 | Exact Match (%) | 52.2 |
| HELM MMLU - College Mathematics | 46 | EM | 50.6 |
| HELM Lite - MedQA | 63.22 | Exact Match (%) | 48.9 |
Interactive version: theaggregate.ai/model?slug=jamba-1-5-mini · How It Works · Data refreshed daily, snapshot 2026-10-20.