Calculate the API cost of transcribing audio using a speech-to-text model.
Speech-to-text providers bill per minute of audio processed. This calculator multiplies your total audio duration by the selected provider's per-minute rate to estimate transcription cost. Rates vary by provider and by features enabled (e.g. speaker diarization, custom vocabulary), so treat these as baseline estimates.
total_cost = audio_minutes × rate_per_minute
No, most providers charge a flat per-minute rate regardless of audio quality, though very poor audio may increase error rates.
Typically yes — most providers bill for the full uploaded duration, including silence, unless you pre-trim the audio.
No, these are base transcription rates. Diarization, custom models, and real-time streaming often carry additional charges.