Automated Transcription Service with Human Accuracy


I just released my first project as a bootstraper.

SpeechText.AI transcription service can accurately transcribe conference calls, interviews, podcasts, lectures, and meeting records in more than 30 different languages and dialects. Our award-winning speech recognition technology achieves a word error rate of 3.8% on the open source LibriSpeech dataset (~1000 hours of clear English speech). SpeechText AI’s speech recognition technology is now almost as accurate as human transcriptionists. Please feel free to create a trial account https:// to see how it works.