Speech-to-text price per hour: Scribe, MAI Streaming, Sume

Speech-to-text vendors quote different units, so the hourly figure is the fair comparison. ElevenLabs lists Scribe v2 at $0.22 per hour and Scribe v2 Realtime at $0.39 per hour on its API pricing page. Microsoft lists MAI-Transcribe-2-Streaming at $0.54 per hour through the end of the year, supporting 60 languages (all read 2026-10-05).

Sume STT is billed at $0.01 per audio minute, which is $0.60 per hour. Each job accepts up to 10 minutes of audio, so an hour is six jobs. The Sume API rate card lists the same figure.

Four-way table

How to read the gap

Sume is the highest of the four per hour. The difference to Scribe v2 batch is $0.38 per hour, or $38.00 over 100 hours. That is the price of an API that already sits next to video, voice and captions tools under one wallet and one key, not a claim about accuracy.

The vendor prices are what the vendors list; they exclude any plan terms, volume tiers or taxes you may have. The MAI price is stated as running through the end of the year, so plan for it to change in January.

Pick the cheaper vendor when transcription is your only workload and you can chunk and bill yourself. Pick Sume when the transcript feeds the next step in the same pipeline and one balance is simpler to forecast.

• Batch file transcription: Scribe v2 is the lowest listed hourly rate.
• Live captions: compare Scribe Realtime with MAI streaming, both priced per hour.
• Pipelines that chain trim, caption and timeline: one wallet removes a vendor.