This study evaluates the Whisper model’s performance for Romanian speech-to-text transcription, investigating how transcription accuracy varies across diverse audio domains. Audio sources, including audiobooks, news ...