A 75-second avatar script as two jobs, 38 s and 37 s: $18.38 at plus
Avatar 1.0 caps one video at 60 seconds. A 75-second script split 38 s + 37 s costs 75 x $0.245 = $18.375 at plus, plus a $0.20 two-minute timeline join.
A 75-second avatar script does not fit in one Sume Avatar 1.0 job, because the accepted window is 4 to 60 seconds. Split it into two jobs of 38 and 37 seconds. At the plus tier that is 38 x $0.245 = $9.31 and 37 x $0.245 = $9.065, for $18.375, and a two-minute timeline render to join them adds $0.20.
Why splitting costs no more per second
Avatar video is billed per second at a flat rate for each tier, so a split does not change the price of the speech itself: 75 x $0.245 = $18.375 either way. The only extra cost is the join. Timeline 1.0 renders cost $0.10 per output minute, rounded up, with a one-minute minimum, so a 75-second output bills two minutes: $0.20.
Check the arithmetic against the single-job ceiling. A 60-second job at plus costs 60 x $0.245 = $14.70, so a 75-second script split as 60 + 15 would cost $14.70 + $3.675 = $18.375, the same as the 38 + 37 split. The even split is chosen for reading rhythm, not for price.
| Step | Seconds | Rate | Cost |
|---|---|---|---|
| Avatar job A | 38 | $0.245 | $9.31 |
| Avatar job B | 37 | $0.245 | $9.065 |
| Timeline render | 75 (billed as 2 min) | $0.10/min | $0.20 |
| Total | $18.575 |
Where to cut the script
Cut at a sentence boundary so neither half ends mid-thought, and keep both halves above the 4-second minimum by a wide margin. Reuse the same avatar_handle and the same scene prompt in both jobs. The docs state the current execution supports one resolved avatar for each final video, so each job renders its own video with the same face rather than switching mid-job.
Join the two videos with the timeline route, or leave them as two separate posts if the platform allows a two-part series. Captions are applied per job when you enable them inline, and a failure in the caption stage is a soft failure: the job can still succeed with a clean video and captions.status set to failed.
- Each job needs its own Idempotency-Key.
- Keep captions identical in style across both jobs, or caption the joined file once with the standalone video captions route.
- Check both clips end on a clean beat before joining.
Check the budget before you start
For longer scripts the same pattern holds: divide by 60 or less, price each piece at the tier rate, add $0.10 per output minute for the join. At standard the same 75 seconds costs 75 x $0.184 = $13.80 plus $0.20; at max it costs 75 x $0.55 = $41.25 plus $0.20.
One more design choice is where the cut falls in the visual story. Because each job renders one resolved avatar and one shared scene, the two halves will look alike, which is what you want for a join. If you want a change of location halfway through, describe the second location in the second job's scene prompt and accept a visible cut. Either way, render the first job, review it, and only then submit the second; a problem in the opening seconds then costs $9.31 rather than $18.375.
Sources
Related posts
More in Sume Avatar 1.0
- Avatar creation is $0.95 once: what reusing one handle saves
Creating an Avatar 1.0 avatar costs a flat $0.95. Reusing the handle across 20 videos of 15 seconds saves $18.05 compared with creating one avatar per video.
- Avatar creation $0.95 over 20 ten-second clips: the per-clip cost
One Sume avatar handle used for 20 ten-second standard clips costs $37.75 in total, or $1.8875 per clip. The arithmetic, with plus and max tiers.
- Avatar handle with a hyphen is rejected: the valid pattern
Sume avatar handles allow letters, digits, periods and underscores, 2 to 30 characters. A hyphen fails the pattern. Check yours in Python first.
- A 37-second AI avatar video: six Sume prices from $6.81 to $21.46
What a 37-second talking avatar video costs on Sume at standard, plus and max, with and without a product image, worked per second.
Written by Sume