ElevenLabs Dubbing v2 plan minutes: 2, 9 or 44, and Sume's steps
ElevenLabs lists Dubbing v2 minutes per plan: 2 on Starter, 9 on Creator, 44 on Pro. Here is what a 10-minute video needs there and how Sume prices the steps.
How many dubbing minutes does each ElevenLabs plan include?
ElevenLabs' Dubbing Studio page lists monthly-billing plans with Dubbing v2 minutes: Free ($0, 10k credits) gives 0.4 minutes, Starter ($6, with $1 for the first month; 10k credits) gives 2, Creator ($22, with $11 for the first month; 30k credits) gives 9, and Pro ($99; 121k credits) gives 44. Extra Dubbing v2 minutes are listed as costing between $4.91 and $2.23 depending on plan.
Dividing price by minutes gives a rough feel: Starter about $3.00 per included minute, Creator about $2.44 and Pro about $2.25. That is arithmetic on the listed figures, not a price ElevenLabs states, and it ignores the first-month discounts and whatever else the credits buy.
What does a 10-minute video need on those plans?
A single 10-minute video exceeds the Starter allowance of 2 minutes and the Creator allowance of 9. Pro's 44 minutes covers it. The page does not say whether minutes are counted per output language, so a four-language release of the same video could mean 40 listed minutes or 10; confirm that before you pick a tier.
The page also says Dubbing v2 conditions on the source performance rather than a transcript, so tone, emotion and delivery carry across languages, and lists 90+ languages and accents, automatic voice cloning and sync-aware translation. It does not state file or speaker limits, and does not mention manual editing or subtitle export, so none of those are assumed here.
| Plan | Price | Credits | Dubbing v2 minutes | Fits one 10-minute video |
|---|---|---|---|---|
| Free | $0 | 10k | 0.4 | No |
| Starter | $6 ($1 first month) | 10k | 2 | No |
| Creator | $22 ($11 first month) | 30k | 9 | No, 1 minute short |
| Pro | $99 | 121k | 44 | Yes |
How does a Sume pipeline charge for the same job?
Sume has no dubbing endpoint, so there is no minutes allowance. You pay per step, per job. Speech-to-text reserves per audio minute and runs once; the listed rate is $0.01 per minute (see the STT pricing post), and a single job takes at most 10 minutes, so a 10-minute video fits one job if you pass duration_seconds. Text-to-speech is priced per character of the translated script and repeats per language. Detach and timeline audio are flat $0.01 per job each.
Because the TTS step scales with script length, not video length, a slow-talking 10-minute video and a fast one cost different amounts. Our step-by-step cost breakdown works through a 1-minute and a 10-minute example, and GET /v1/catalog always has the live rates.
What do you lose and gain going step by step?
You lose the audio-to-audio conditioning: a transcript-based pipeline cannot carry a whispered aside or a laugh the way a model that listens to the performance can. You also lose one-click voice cloning of the speaker; you choose a TTS voice. You gain separate review points, because the transcript and the translation are text you can check before any voice is made, and a bad sentence is a one-line retry.
The pipeline in short: audio detach for the track, STT with sentence segmentation, your translation step, TTS per sentence with language set, then timeline audio to join the lines or audio.parts[] in a Timeline 1.0 render. The Dubbing v2 project API post compares the workflows.
Which one should you buy?
One more check applies to both routes: rights. Dubbing a voice into another language, or cloning it, can require the speaker's consent, and some platforms ask for a disclosure when audio is synthetic. Neither a plan page nor an API job settles that for you, so get the permission in writing before you scale up a catalogue of dubbed videos.
Count the minutes you will dub in a month, multiply by languages if that is how the plan counts, and compare against Pro's 44. If the volume is a few videos and performance matters, a plan with included minutes is simpler. If you dub constantly, want to control each step, or already have the translated script, per-step pricing is easier to forecast per sentence. Either way, run a single 60-second test in your target language and have a native speaker listen before you commit a month of budget.
Sources
Related posts
More in Pricing
- ElevenLabs v4 after October 12: $0.08 per 1,000 characters vs Sume
ElevenLabs v4 returns to $0.08 per 1,000 characters after the October 12 promo, Turbo to $0.04. Sume TTS lists $0.0475 per 1,000. Script costs side by side.
- Estimate a Sume image batch cost from endpoint pricing lines in Python
Estimate an image batch before you run it: read cost_usd from the model's endpoint pricing line and multiply by n. A short Python script, with a caveat.
- Charged for a failed Seedance or Kling job? Reserve, capture, refund
How Sume bills a Seedance or Kling video: a hold at submit, capture on success, refund on failure, and cancel only before generation starts.
- fal H100 at $2.49/hour vs per-output media API pricing: which fits
fal lists an H100 at $2.49 per hour and video APIs from $0.025 to $0.14 per second. When hourly GPU pricing beats paying per output, and when it does not.
Written by Sume