A 10-minute avatar module on Sume: ten jobs, priced by tier
Sume avatar videos run 4 to 60 seconds, so a 10-minute module is ten jobs and a join. Standard comes to $110.40 and plus to $147.00, by arithmetic from list.
On Sume a 10-minute avatar module is ten avatar video jobs of 60 seconds each, joined in Timeline. At the listed per-second rates that is $110.40 on the standard tier, $147.00 on plus and $330.00 on max, plus about $1.11 to join and $0.95 once if you create a new avatar. HeyGen's Avatar V page, by comparison, says there is no cap on generated length and lists no price, so the two are not directly comparable.
What do the tiers cost?
Avatar video is billed by estimated duration, which must be 4 to 60 seconds per job. The rates below are for jobs without a product image.
| Tier | Rate per second | 600 seconds |
|---|---|---|
| standard | $0.184 | $110.40 |
| plus (default) | $0.245 | $147.00 |
| max | $0.55 | $330.00 |
What does the join add?
Each part has its own audio. You can detach each track with Audio detach at $0.01 per job, which is $0.10 for ten, then concatenate with Timeline audio at $0.01, then render the ten clips in Timeline at $0.10 per output minute, so $1.00 for 10 minutes. That is about $1.11. Timeline sources must be media.sume.com files, so confirm each avatar result URL is on that host before joining.
What can differ between the parts?
Each job is generated separately. Use the same avatar handle, aspect ratio and tier in every job so the parts match. Preview a part with the tier-independent preview first. Sume does not state that the parts are free of identity drift, and you should check them yourself; the drift check post shows how with Video inspect stills.
What does HeyGen say about length?
HeyGen's Avatar V page says there is no cap on generated length and claims no identity drift across a 30 second clip or a 10-minute module. Those are HeyGen's claims about its own product. The page shows no price.
How should I plan the script?
Write the script in ten blocks that each read in under 60 seconds, with a clean sentence break at the end of each. A block that runs long is cut into two jobs, which raises the count. Preview one block on the cheapest tier, check the voice and framing, then submit the rest. Estimated duration drives the price, so shorter blocks cost less.
Sources
Related posts
More in Use cases
- Tennessee ELVIS Act and AI voice: a checklist for TTS projects
Tennessee's ELVIS Act, signed 21 March 2024, extends protection to voice. A consent checklist for AI voiceover, using the governor's page and Sume's TTS docs.
- Text edited in an ad image: a dimension check before you ship
After an AI text replacement, compare output and source dimensions in Pillow. Stretched or recropped edits are caught in one assert, before the ad goes live.
- Text to speech from a PDF with an API: extract, chunk, narrate
Sume's TTS takes text, not PDF files. Extract the text, split it under 20,000 characters, narrate each chunk, then join the audio. Costs and limits included.
- Thanksgiving recipe video: step cards with caption cues
Make a 40-second Thanksgiving recipe video from five phone shots: join them in Timeline 1.0, burn numbered step cards with caption cues, add a music bed.
Written by Sume