SEA-HELM (Vietnamese) - Sentiment Analysis: leaderboard
Sentiment classification of Vietnamese text. Normalized 0-100 score from AI Singapore's SEA-HELM leaderboard: (score - random baseline) / (maximum - baseline), floored at 0 and averaged over eight runs; higher is better.
Metric: Normalized Score. Source: leaderboard.sea-lion.ai. Saturation forecast: Estimated already saturated. 58 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 3 12B (IT) | 65.86 | #672 |
| 2 | Mistral Small 4 | 64.67 | #522 |
| 3 | Gemma 3 4B (IT) | 64.57 | #997 |
| 4 | Qwen 3 VL 4B Instruct | 64.41 | #510 |
| 5 | Llama-SEA-LION-v3-8B-IT | 63.88 | #800 |
| 6 | OLMo 3.1 32B Instruct | 62.58 | #755 |
| 7 | Gemma 3 27B (IT) | 62.55 | #528 |
| 8 | Qwen 3.5 27B (Reasoning) | 61.99 | #192 (Qwen 3.5 27B) |
| 9 | Llama-SEA-LION-v3-70B-IT | 61.52 | #407 |
| 10 | Qwen 3.6 27B (Reasoning) | 61.32 | #182 (Qwen 3.6 27B) |
| 11 | Qwen 3.5 122B A10B (Reasoning) | 61.18 | #184 (Qwen 3.5 122B A10B) |
| 12 | MERaLiON-2-10B | 61.11 | #600 |
| 13 | Qwen 3 VL 32B Instruct | 61.1 | #290 |
| 14 | Qwen 3.5 35B A3B (Reasoning) | 61.06 | #268 (Qwen 3.5 35B A3B) |
| 15 | Qwen 3 VL 8B Instruct | 60.93 | #419 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=sea-helm-vietnamese-sentiment-analysis · How It Works · Data refreshed daily, snapshot 2026-11-05.