Sonix $10 an hour vs Sume's $0.01-a-minute transcript

Sonix lists Pay As You Go at $10 an hour, which is about $0.167 a minute. Sume prices the transcript on a video inspect at $0.01 per audio minute, which is $0.60 an hour. That is a 16-fold gap on the line item, but the two are not the same product, and Sume adds an inspect compute charge on top. This post sets both sides out so you can decide with real numbers.

What the Sonix page says

The Sonix pricing page lists a single-user Pay As You Go plan at $10 per hour and says billing is prorated down to the nearest second. It also lists three subscriptions (Core, Advanced, Pro), says extra hours on any subscription cost $10 an hour, and offers new accounts 30 free minutes with no card.

The part of the Sume price that is not in the table

On Sume the transcript is an option on a video inspect job. The docs say each inspect reserves its compute ceiling and captures real container seconds at the Modal list rate times 1.25, plus the platform fee, and that the transcript adds its per-minute rate to that reservation. So the all-in number for an hour of audio is above $0.60. The docs do not give a flat all-in figure for transcription alone, so measure it: run one real file and read the cost on the job.

Limits to check before you switch

• The clip must live on media.sume.com. The API does not fetch from the open internet, so you import first with POST /v1/media-imports.
• Without a duration_seconds hint Sume reserves one minute; the largest hint the docs allow is 600 seconds. Read the video inspect page for how longer files behave before you plan a 3-hour batch.
• Writes need an Idempotency-Key header.

Steps: price one file on Sume

• Import the file with POST /v1/media-imports and an Idempotency-Key.
• Submit a video inspect with transcribe set to true and your duration_seconds hint.
• Poll the job, read usage.cost, and divide by the audio minutes.
• Compare that per-minute figure to $0.1667.

curl -X POST https://api.sume.com/v1/video-inspect \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: transcript-price-001" \
  -d '{"video_url":"https://media.sume.com/artifacts/artf_demo/talk.mp4","transcribe":true,"duration_seconds":300}'

What Sume does not do

Sume has no transcript editor, no team seats and no subscription bundle of hours. It returns text and optional sentence segments through an API. If your people live in an editor and share transcripts, Sonix is built for that; if a program needs the words, Sume is the cheaper line.

A worked monthly example

Take a podcast network with 12 hours of new audio a month. On Sonix Pay As You Go that is $120. On Sume the transcript line is 720 minutes at $0.01, or $7.20, and the inspect compute comes on top. The compute figure depends on container seconds, so run three representative files and average the usage.cost you read, then add the $7.20. If the all-in number comes out under $120 by a wide margin, the pipeline wins on cost; if you also need an editor, factor in the engineering time you would spend building one.

Which to pick

Pick Sonix when a person edits and shares transcripts and your volume is occasional. Pick Sume when transcripts are an automated step inside a video pipeline that already imports clips. At 10 hours a month the list-rate difference is $100 against $6 plus compute, which is worth measuring once.