AnyTranscription greatly improves the speed and accuracy of transcription by expanding its speech and word corpus resources and utilizing a weighted finite-state transducer (WFST) decoding network base.