# Speech To Text

Ranked by the Vioscale composite score: a confidence-weighted blend of independent signals, not user reviews.

| # | Software | Score | Confidence | Vendor |
|--:|---|--:|---|---|
| 1 | [Speechmatics](https://www.vioscale.ai/software/speechmatics) | 81.4 | 55% (medium) | - |
| 2 | [AssemblyAI](https://www.vioscale.ai/software/assemblyai) | 80.2 | 60% (medium) | - |
| 3 | [whisper.cpp](https://www.vioscale.ai/software/whisper-cpp) | 68.8 | 24% (low) | - |
| 4 | [Deepgram](https://www.vioscale.ai/software/deepgram) | 68.3 | 43% (low) | - |
| 5 | [Amazon Transcribe](https://www.vioscale.ai/software/amazon-transcribe) | 65.3 | 34% (low) | - |
| 6 | [faster-whisper](https://www.vioscale.ai/software/faster-whisper) | 38.9 | 28% (low) | - |
| 7 | [Vosk](https://www.vioscale.ai/software/vosk) | 37.5 | 28% (low) | - |
| 8 | [SpeechBrain](https://www.vioscale.ai/software/speechbrain) | 26.8 | 23% (low) | - |
| 9 | [Kaldi](https://www.vioscale.ai/software/kaldi) | 17.8 | 15% (low) | - |

---
*Source: Vioscale (https://www.vioscale.ai/categories/speech-to-text).*