Mux premium captions 500 units a minute vs Sume $0.20 per job

Mux Robots generate-premium-captions consumes 500 units per minute of video, and Mux prices 1,000 units at $0.01, so the workflow costs $0.005 per minute once the 100,000 free monthly units are used. Sume's video captions reserves and captures $0.20 per accepted job for videos up to 60 seconds, but it does a different thing: it burns styled captions into a new video.

Those two prices are not interchangeable, and this page explains why, from Mux's pricing page and the Sume docs.

What does Mux actually charge for premium captions?

The Mux pricing page states that the first 100,000 units each month are free and that 1,000 units equal $0.01. The generate-premium-captions workflow uses 500 units per minute, so the free allowance covers 200 minutes of video each month, then $0.005 per minute.

Two extras are worth reading before you budget. The page says the workflow may create temporary audio-only static renditions, which incur small Mux Video charges billed separately. And Mux's changelog notes that Mux reduced the premium captions price on September 17, 2026 when the workflow left experimental status; the pricing page above is the current figure, so use it rather than older quotes.

What does Sume charge, and for what?

Standalone video captions is POST /v1/video-captions with a public HTTPS video_url. It returns a job-backed captioned video. The docs say each accepted job reserves and captures $0.20 for videos up to 60 seconds under the current fixed estimate, and tell you to confirm live pricing in GET /v1/catalog.

Sume's price is per job, not per minute, and it includes the render that burns the caption styles (slam, punch, tiktok-green, and Hangul styles such as black-outline). Docs for videos longer than 60 seconds do not state a different rate, so check the catalog before running long files.

Are they the same product?

No. Mux's workflow produces captions for a Mux Video asset as a text track for playback; the Robots guide groups it under accessibility. Sume's job returns a new MP4 with the captions drawn onto the pixels, and also supports authored cues that skip speech-to-text for silent clips. If you need a sidecar caption track, Sume does not produce one: the docs state SRT uploads are unsupported as input, and the result is a captioned video.

So the cheaper per-minute number only wins if a text track is what you need. If you need burned-in captions on a clip for TikTok or Reels, the comparison is Mux plus your own render step versus one Sume call.

How do you try Sume captions cheaply?

Submit with an Idempotency-Key, so a retry never bills twice, and use source_caption_id to restyle without transcribing again. Billing is unchanged for a restyle, because it is still a render.

curl -X POST https://api.sume.com/v1/video-captions \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: caption-demo-001" \
  -d '{
    "video_url": "https://media.sume.com/artifacts/example/clean.mp4",
    "style": "slam"
  }'

What does a month of captions cost in each?

Worked example from the published numbers only. Suppose you caption 400 minutes of video in a month. On Mux, 400 minutes at 500 units is 200,000 units; the first 100,000 are free, and the remaining 100,000 units cost $1.00 at 1,000 units per $0.01, plus any separate Mux Video charges. On Sume, if those 400 minutes are 400 clips of one minute or less, that is 400 jobs at $0.20, or $80.00.

The gap is large, but you are buying different things: Mux's figure is for captions on assets already in Mux, while Sume's includes the burned-in render. If your clips are longer than 60 seconds, the Sume docs do not state a different rate, so confirm in GET /v1/catalog first. Do not extrapolate from this example to long files.

• Mux: 500 units per video minute, 1,000 units per $0.01.
• Sume: $0.20 per accepted job for videos up to 60 seconds.
• Both prices were read on 2026-10-02 and can change.