FMSU-Bench - Speaking Rate (Chinese): leaderboard

FMSU-Bench (2026) tests fine-grained speech understanding in Chinese and English with multiple-choice questions whose distractors include subtle acoustic confusions and text-only semantic traps, plus a tagged transcription task. This board is accuracy on identifying the speaking-rate characteristics in the Chinese subset; percent, higher is better.

Metric: Accuracy (%) of identifying the speaking-rate characteristics (SR), on the FMSU-Bench Chinese subset (500 to 1,000 manually verified items per task, from audiovisual sources), multiple-choice questions whose distractors include fine-grained acoustic confusions and text-only semantic traps; higher is better. Source: arxiv.org. Saturation forecast: Around February 2028. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 3 Omni 30B A3B75.6#847
2Qwen2.5-Omni-7B75.4#1884
3Gemini 3 Flash74.7#175
4Gemini 3.1 Pro (Preview)69.8#99
5Gemini 2.5 Flash68.1#463
6MiMo-Audio58.6#931
7Kimi-Audio37.9#1735
8FMSU-Bench Step-Audio 2 (checkpoint unspecified)34.7
9Qwen2-Audio (FMSU-Bench checkpoint unspecified)22.3
10Audio Flamingo 321.4

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=fmsu-bench-speaking-rate-chinese · How It Works · Data refreshed daily, snapshot 2026-11-21.