Tongyi DeepResearch 30B A3B: benchmark results
Provider: Other. Released 2025-09-16. Access: Open.
Unified ELO 1687 ± 28, rank #351 of 2656 rated models, from 30 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| TurkBench - Bias | 84.2 | Accuracy (%) | 75.7 |
| SnakeBench | 25.7 | TrueSkill Rating | 69.6 |
| TurkBench - NLI | 75.69 | Accuracy (%) | 60.8 |
| TurkBench - Toxicity | 98.28 | LLM-as-a-judge score (0-100) | 59.5 |
| TurkBench - Metaphors and Idioms | 68 | Accuracy (%) | 58.1 |
| TurkBench - Math | 29 | Accuracy (%) | 54.1 |
| TurkBench - Sentiment Analysis | 45.99 | Accuracy (%) | 54.1 |
| TurkBench - Turkish Vocabulary | 83 | Accuracy (%) | 51.4 |
| TurkBench - Named Entity Recognition | 55.09 | Accuracy (%) | 45.9 |
| TurkBench - Faithfulness | 85.57 | LLM-as-a-judge score (0-100) | 40.5 |
| TurkBench - Summarization | 72.92 | LLM-as-a-judge score (0-100) | 40.5 |
| TurkBench - Turkish General Knowledge | 64.58 | Accuracy (%) | 40.5 |
Interactive version: theaggregate.ai/model?slug=tongyi-deepresearch-30b-a3b · How It Works · Data refreshed daily, snapshot 2026-09-19.