What does $1 of Sume balance buy? Seconds, tracks, images, minutes

$1 buys 100 minutes of speech-to-text, 21,052 characters of voice, 8 music tracks, 16 seconds of Wan 480p video, 10 Nano Banana 2 images or 200 Soul stills.

4 min readSume
All posts

One dollar of Sume balance buys 100 minutes of transcription, 21,052 characters of text to speech, 8 music tracks, 16 seconds of Wan 3.0 video at 480p, 10 Nano Banana 2 images at 1K, or 200 Higgsfield Soul images at 720p. It is a small unit that shows how different the SKUs are.

What $1 buys (read 2026-10-04)
JobSume rate$1 buys
Speech-to-text$0.01 per minute100 minutes
Text to speech$0.0475 per 1,000 chars21,052 characters
Music track$0.1258 tracks
Wan 3.0 480p video$0.0625 per second16 seconds
Nano Banana 2, 1K image$0.1010 images
Higgsfield Soul 720p image$0.005200 images
Background removal$0.022544 images
Captions job$0.205 jobs
Timeline render$0.10 per output minute10 minutes

Where the rates live

Every rate is on the rate card and in GET /v1/catalog, described in the API reference. Image prices by model are in the Images docs; video by model in the Video Router docs.

These are one-job rates

Real projects mix jobs: a short film is video plus voice plus music plus a render. Add each line item at its own rate; do not scale one row of this table into a whole production.

Recompute it in Python

One dollar divided by each rate. The numbers are the ones in the table above.

rates = {'stt minutes': 0.01, 'music tracks': 0.125, 'wan 480p seconds': 0.0625, 'soul images': 0.005}
for name, rate in rates.items():
    print(name, round(1 / rate, 1))

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume