Nemotron-SEA-LION-v5-E4B: benchmark results
AI Singapore's SEA-LION v5 E4B model (Nov 18, 2026, MIT), post-trained on NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 (8.0B total, 4B effective). Provider: AI Singapore. Released 2026-11-18. Access: Open.
Aggregate Rating 1652 ± 36, rank #561 of 10524 rated models, from 132 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SEA-HELM (Filipino) - Belebele QA | 89.17 | Normalized Score | 97.3 |
| SEA-HELM (Indonesian) - Sentiment Analysis | 82.34 | Normalized Score | 97.3 |
| SEA-HELM (Filipino) - SEA-IFEval | 93.44 | Normalized Score | 95.9 |
| SEA-HELM (Malay) - SEA-IFEval | 89.78 | Normalized Score | 95.9 |
| SEA-HELM (Filipino) - NLG | 57.74 | Normalized Score | 94.6 |
| SEA-HELM (Filipino) - NLU | 76.84 | Normalized Score | 94.6 |
| SEA-HELM (Filipino) - Summarization | 25.33 | Normalized Score | 94.6 |
| SEA-HELM (Thai) - SEA-IFEval | 95.07 | Normalized Score | 94.6 |
| SEA-HELM (Burmese) - Sentiment Analysis | 47.97 | Normalized Score | 93.2 |
| SEA-HELM (Indonesian) - Summarization | 18.06 | Normalized Score | 91.9 |
| SEA-HELM (English) - IFEval | 93.91 | Normalized Score | 91.7 |
| SEA-HELM (Thai) - Thai Exam | 76.24 | Normalized Score | 90.5 |
Interactive version: theaggregate.ai/model?slug=nemotron-sea-lion-v5-e4b · How It Works · Data refreshed daily, snapshot 2026-11-22.