List your company

Temporal Grounding (Word Alignment) on Librispeech

7.7Accuracy (20ms tolerance)Gemini 2.5 Flash
-0.2041.8483.95.952Feb 10, 2026

Evaluation results

MethodLinks
Gemini 2.5 FlashInference Mode=Zero-shot2026.027.713.80.28
Voxtral 3BInference Mode=Zero-shot2026.020.81.12.52
Qwen2.5 7BInference Mode=Zero-shot2026.020.71.12.17
Audio Flamingo 3Inference Mode=Zero-shot2026.020.61.41.82
GPT-4o AudioInference Mode=Zero-shot2026.020.30.52.42
Voxtral 24BInference Mode=Zero-shot2026.020.20.52.12
Qwen2.5 3BInference Mode=Zero-shot2026.020.10.22.59