Automatic Speech Recognition on SAP rated (test)
13.38WERQwen3-ASR-1.7B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-ASR-1.7BPromptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 13.38 | |
| Qwen3-Omni-30B‡Promptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 13.77 | |
| Voxtral-Small-24BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 13.88 | |
| Cohere TranscribePromptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 14.48 | |
| Voxtral-Mini-3BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 15.31 | |
| Phi-4 MultimodalPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 15.77 | |
| Parakeet-TDT-1.1BPromptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 15.85 | |
| Qwen3-Thinking-30BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 16.5 | |
| Whisper-large-v3Promptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 17.1 | |
| MiniCPM-o 4.5Promptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 18.12 | |
| MiniCPM-o ThinkPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 18.41 | |
| Qwen2-Audio-7BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 22.22 | |
| Ultravox-Nemo 12BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 31.23 | |
| Gemma-4-4BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 31.61 | |
| Gemma-4-2BPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 31.96 | |
| Gemma-4-4B ThinkPromptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 32.15 | |
| Wav2Vec2-RobustPromptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 32.64 | |
| Wav2Vec2Promptable=No, Evaluation Mode=Zero-shot baseline2026.05 | 39.3 | |
| Audio Flamingo 3∗Promptable=Yes, Evaluation Mode=Zero-shot baseline2026.05 | 49.55 |