CoDeTT (Chinese, No History): leaderboard
Metric: Macro accuracy (%) over the four turn-taking actions (Maintain, Stop & Listen, Takeover, Dismiss) on the Chinese CoDeTT instances, with no earlier dialogue rounds; audio input, a constant answer scores 25; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 80.83 | #77 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=codett-chinese-no-history · How It Works · Data refreshed daily, snapshot 2026-10-11.