HeyGen Avatar V on HiAPI costs $0.15 per second of output video: $1.50 for a 10-second clip, the same rate as Avatar IV. The rate is identical at 720p, 1080p and 4k, so your budget follows the finished video's length.
For a first test, use Daniel's script below. The existing sample runs about eight seconds, a budget of about $1.20; a new task settles to its actual length. Generate with this prompt → Manually choose Brandon Office Sitting Front, Dexter - Professional, 16:9, 1080p and MP4.
How much does each video cost?
| Video length | Budget per video |
|---|---|
| 5 seconds | $0.75 |
| 10 seconds | $1.50 |
| 15 seconds | $2.25 |
| 30 seconds | $4.50 |
| 60 seconds | $9.00 |
| 120 seconds | $18.00 |
HiAPI holds an estimated amount when you create a task, then settles it against the actual output seconds when the video finishes. The difference is refunded or topped up automatically; failed tasks are refunded in full.
There is no duration field to set. Script length and speaking pace determine how long a script-driven video runs. With an audio track, the recording determines the length. Estimate the read-aloud time before submitting a long script.
$10 buys about 66 seconds of video. Use a short line to choose your presenter and voice, then move on to the full script. Generate with this prompt →.
What does a month of videos cost?
Multiply videos per month by seconds per video by $0.15:
| Videos per month | 10 seconds each | 30 seconds each | 60 seconds each |
|---|---|---|---|
| 20 | $30 | $90 | $180 |
| 100 | $150 | $450 | $900 |
| 300 | $450 | $1,350 | $2,700 |
For example, 300 ten-second social clips total 3,000 seconds and cost $450. Twenty one-minute lessons total 1,200 seconds and cost $180. Budget in total seconds first, then split those seconds into the clips your publishing schedule needs.
Tighter scripts make a measurable difference: cutting five seconds from each of 100 videos saves $75. Put the message first, and remove repeated introductions before sending a batch. Generate with this prompt → Check the presenter and voice with one line before preparing the batch; select the remaining settings manually.
Do 4k, captions or transparent backgrounds cost extra?
720p, 1080p and 4k share the same per-second rate. Aspect ratio, captions and output format do not have separate price tiers either. Choose 1080p for a web video or 4k when your workflow calls for a higher resolution. The API defaults to 720p, so set resolution explicitly.
Enable caption to receive a separate SRT file you can import into an editor. For a transparent asset, use remove_background=true and output_format=webm. Try a short script first and check the selected presenter's exported background before building the rest of your edit around it.
This Daniel briefing uses Brandon Office Sitting Front and Dexter - Professional. The request selected 1080p and 16:9; the web video is 1280×720 and runs about eight seconds:
Hi, I'm Daniel. Send me your quarterly numbers, and I'll walk your team through what changed and what to do next.
Generate with this prompt → Only the script is filled in. Choose the presenter, voice, 16:9, 1080p and MP4 yourself. The API tutorial covers audio input, captions and background settings.
Avatar IV or Avatar V: which should you budget for?
| Choice | Avatar IV | Avatar V |
|---|---|---|
| Presenter | Your uploaded portrait | 497 preset avatars |
| Resolution | 720p / 1080p | 720p / 1080p / 4k |
| Output options | MP4 | MP4, transparent WebM, separate SRT |
| Price per second | $0.15 | $0.15 |
Use IV when a specific person in your photo needs to appear on screen. Choose V when a preset presenter fits the job, or you need captions, transparent assets or 4k. The rate is the same, so decide based on the source of your presenter and the output your workflow needs. See the Avatar V parameter reference for the available settings.
How do you avoid paying for extra runs?
- Test the setup with a short line. Confirm the presenter, voice and background before rendering a long script.
- Continue querying an existing task. Every POST creates a new task. Once you have a
taskId, use it to check progress; if the submission response is lost, check your task history before submitting again. - Save successful outputs promptly. Download the video and caption files before
expireAt, then keep them in your own asset library.
For a portrait announcement with SRT captions, follow the vertical example in the API tutorial.
FAQ
Does using my own audio change the rate? No. Audio-driven video is also $0.15 per output second. Input audio can run up to 600 seconds.
Will the estimated charge adjust automatically? Yes. The final charge uses actual output seconds, with the difference settled automatically. Failed tasks receive a full refund.
Does a lower resolution reduce the bill? No. The per-second rate is the same. Reducing total output by ten seconds reduces the budget by $1.50.
Start with the business script above. Generate with this prompt → The script is filled in; choose your presenter, voice, aspect ratio and output settings. To automate production, follow the Avatar V API tutorial; for the accepted input values, browse the Avatar V input reference.
