ElevenLabs audio API prices next to the Sume audio jobs that overlap
ElevenLabs lists per-minute prices for dubbing, music, isolator and speech-to-text. Sume documents flat per-job prices for detach, concat, captions and STT.
The prices are quoted in different units, so compare tasks rather than numbers. ElevenLabs lists most audio services per minute or per 1,000 characters; Sume documents flat prices per job for audio detach, audio concat or split, and video captions, plus a per-audio-minute rate for speech-to-text. The docs list no dubbing or voice isolation job, so those rows have no Sume counterpart.
This is a reference sheet. Verify the live numbers on each vendor's page before you budget.
ElevenLabs as listed
The ElevenLabs API pricing page lists the following. The v4 price shows a launch discount that ends on October 12.
| Service | Listed price |
|---|---|
| Eleven v4 text to speech | $0.022 per 1K characters until Oct 12, regular $0.08 |
| Eleven v4 Turbo | $0.011 per 1K characters, regular $0.04 |
| Eleven v3 / Multilingual v2 | $0.08 per 1K characters |
| Music | $0.15 per minute |
| Scribe v2 | $0.22 per hour |
| Scribe v2 Realtime | $0.39 per hour |
| Dubbing v1 | $0.33 per minute watermarked, $0.50 clean |
| Dubbing v2 | $2.20 per minute |
| Voice Isolator, Voice Changer, Sound Effects | $0.12 per minute each |
Sume as documented
The Sume docs state flat prices for utility jobs: audio detach is $0.01 per job, timeline audio concat or split is $0.01 flat per job, video captions is $0.20 for videos up to 60 seconds, and Timeline 1.0 renders at $0.10 per output minute rounded up. Speech-to-text has a public rate of $0.01 per audio minute in the video inspect docs. This post states no Sume TTS price, so check the live catalog for it.
| Task | Sume surface | Documented price |
|---|---|---|
| Extract audio from a video | Audio detach | $0.01 per job |
| Join or slice audio files | Timeline audio | $0.01 flat per job |
| Burned-in captions | Video captions | $0.20 up to 60 seconds |
| Render with voice and music | Timeline 1.0 | $0.10 per output minute, rounded up |
| Transcribe speech | STT 1.0 | $0.01 per audio minute |
How to compare
Pick the task and price it both ways with your own volume. Transcribing a 10-minute clip is 10 minutes on a per-minute rate; a 20-second caption job is one flat $0.20 job. Dubbing, isolation and sound-effects generation have no Sume job at all, so the choice there is not price but whether you do the step outside Sume.
- Compute per task, not per headline rate.
- Include the discount end date when a launch price is shown.
- Add the cost of moving files between tools.
- Re-read both pricing pages on the day you decide.
Sources
Related posts
More in Pricing
- ElevenLabs Eleven v4 72% off until Oct 12: the per-character math
ElevenLabs lists Eleven v4 at $0.022 per 1K characters until Oct 12, then $0.08; v4 Turbo is $0.011, then $0.04. Costs for 10,000 and 100,000 characters.
- ElevenLabs v4 cost of a 500, 1,500 or 5,000-character script
At ElevenLabs list prices a 1,500-character script costs $0.033 on v4 until Oct 12, then $0.12; on v4 Turbo $0.0165, then $0.06. Table for 500 to 5,000 chars.
- Fabric talking clip cost by audio length: 480p vs 720p table
VEED Fabric 1.0 on Sume bills per audio second: $0.10 at 480p, $0.1875 at 720p, up to 300 seconds. Cost of 15, 30, 60, 120 and 300 second clips.
- $50 AI video budget: usable 8-second clips by takes per keeper
$50 buys 50 clips of 8 s on Wan 3.0 at one take, 16 at three; Seedance 2.5 buys 10 or 3. The take rate matters as much as the model. Table inside.
Written by Sume