Qwen-SEA-LION-v5-9B: benchmark results

AI Singapore's SEA-LION v5 9B multimodal model (Nov 18, 2026, MIT), post-trained on Qwen3.5-9B. Provider: AI Singapore. Released 2026-11-18. Access: Open.

Aggregate Rating 1627 ± 35, rank #708 of 10524 rated models, from 132 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SEA-HELM (Filipino) - SEA-Safeguard77.85Normalized Score91.9
SEA-HELM (Thai) - Safety59.77Normalized Score91.9
SEA-HELM (Tamil) - SEA-Safeguard82.27Normalized Score90.5
SEA-HELM (Thai) - Toxicity Detection46.07Normalized Score90.5
SEA-HELM (Vietnamese) - Question Answering80.35Normalized Score83.8
SEA-HELM (Thai) - NLU63.42Normalized Score79.7
SEA-HELM (Thai) - SEA-Safeguard73.47Normalized Score79.7
SEA-HELM (Thai) - Sentiment Analysis40.36Normalized Score79.7
SEA-HELM (Thai) - Question Answering86.48Normalized Score78.4
SEA-HELM (English) - MathArena40.32Normalized Score78.3
SEA-HELM (English) - AA-LCR66.71Normalized Score76.7
SEA-HELM (Burmese) - Sentiment Analysis43.81Normalized Score75.7

Interactive version: theaggregate.ai/model?slug=qwen-sea-lion-v5-9b · How It Works · Data refreshed daily, snapshot 2026-11-22.