Speech Recognition on Aura-STT-BenchLF MSA
16.3WER (%)Aura-STT-LF
Evaluation Results
| Method | Links | |
|---|---|---|
| Aura-STT-LFParams=479M2026.03 | 16.3 | |
| Aura-STT-LF2026.03 | 16.66 | |
| Whisper-large-v3Params=1.55B2026.03 | 20.42 | |
| Nvidia-Conformer-CTCParams=121M2026.03 | 22.81 | |
| Aura-STT*Params=446M, configuration=VAD + Aura-STT (short-form), beam-size=1 (greedy search)2026.03 | 24.51 | |
| Gemini 2.5 Flash2026.03 | 26.79 | |
| Whisper-v32026.03 | 28.08 | |
| OWSM-CTC-v4Params=1.0B2026.03 | 64.87 |