Whisper API cost per minute, and what an hour costs

OpenAI's pricing page lists Whisper and gpt-4o-transcribe at $0.006 per minute and gpt-4o-mini-transcribe at $0.003 per minute. That is $0.36 and $0.18 for an hour of audio. Sume STT 1.0 is $0.01 per audio minute, which is $0.60 for an hour.

OpenAI numbers are as written on its pricing page, read 2026-09-30. Sume's rate is from the Video inspect docs. This post compares prices only, not accuracy or features.

What do one hour and ten hours cost on each line?

Multiply the per-minute price by 60 for an hour, or by 600 for ten hours.

Why is the Sume rate shaped like this?

In the code the Sume rate is the provider list price of $0.008 per minute at the standard margin of 1.25, which gives the public per-minute figure.

What does a Sume STT request return?

Completed results expose text, language fields when available, and words[] word-level timings in seconds from the start of the audio. Add language_code as a hint, or omit it for auto-detect. See speech-to-text word timestamps.

How long can one request be?

Omit duration_seconds and Sume reserves one minute; the maximum is 10 minutes per request. A one-hour recording therefore needs several requests; the steps are in transcribe long audio files.