How big is a TTS mp3 per minute at 32 to 192 kbps?

A constant-bit-rate mp3 takes bit_rate / 8 * 60 bytes per minute. For Sume TTS that gives 240 KB per minute at 32 kbps, 480 KB at 64, 720 KB at 96, 960 KB at 128 and 1.44 MB at 192. Real files run a little larger because of tags and headers. These are the five mp3 bit rates the TTS 1.0 contract accepts, and 128 kbps at 44,100 Hz is the default (API reference).

Size per minute

The table is arithmetic from the bit rate, not a measurement of any one file.

Which rate for a voiceover

Speech is a narrow signal, but the voiceover is rarely the final file. It is an input to a video render or a join, so keep quality high until the last step. Two facts from the docs push the same way.

• mp3 carries encoder priming padding, and Sume documents that joined mp3 parts can show small gaps. The timeline audio docs default to wav for that reason.
• If you only need a preview for a client to approve, 64 kbps is half the size of the 128 default and the saving is real when you send hundreds of files.
• If the next step is a render or a join, ask for wav at the sample rate you will use, and skip the mp3 round trip.

Set it in the request

Put the rate in output_format. The bit_rate field is for mp3 only, and the API reference says it is required when you override the mp3 defaults. Cost does not change with the file type: the price is $0.0475 per 1,000 characters either way, so only storage and transfer move.

For comparison, Microsoft quotes MAI-Voice-2.1 at $22 per million characters and the Flash tier at $15 per million (Microsoft AI, read 2026-10-04). Those are list prices per character, and neither figure depends on file size.

{
  "transcript": "Your script here.",
  "avatar_handle": "your-voice-handle",
  "output_format": {"container": "mp3", "sample_rate": 44100, "bit_rate": 64000}
}

Check before you ship

Run one 60-second take at 64 and 128 kbps, listen on a phone speaker, and keep the lower rate only if you cannot hear a difference. The arithmetic above tells you what you save: at 64 kbps one minute is 960 KB smaller than at 192.