Skip to main content
For the best AI voice cloning results, make your dynamic videos at least 10 seconds long. Videos under 5 seconds often fail because the AI lacks enough voice input. To keep messages longer while saving minutes, record a short dynamic portion and extend it with a combined (static) video, which doesn’t consume dynamic video minutes.

​
Key takeaways

  • Aim for a minimum of 10 seconds so the AI has enough audio to generate a natural-sounding voice; under 5 seconds often fails completely.
  • The entire dynamic video length counts toward your DVM usage — even if the personalized variable is only a one-second greeting.
  • To save minutes, record only the dynamic, personalized portion (≥ 10s) and add a combined video for the rest, which does not use dynamic video minutes.

​
🧠 Guidelines for Creating Effective Dynamic Videos (and How Minute Usage Works)

When creating dynamic videos in Sendspark using AI voice cloning, it’s important to provide enough input for the AI to work with.
If the video is too short (under 5 seconds), the AI will likely fail to generate a natural-sounding voice because there simply isn’t enough voice input.

​
🎯 Why We Recommend a Minimum of 10 Seconds

To ensure the best possible result, we recommend that dynamic videos be at least 10 seconds long.
This gives the AI enough audio to generate a natural-sounding voice and keeps the video from sounding abrupt or “robotic.” Videos shorter than 10 seconds are prone to quality issues, and videos under 5 seconds often fail completely.

​
🧩 Want to Save Minutes? Use Combined Videos

If you’re trying to save DVM minutes, here’s the recommended approach: Suggested Workflow
  1. Record the dynamic, personalized portion of the video (≥ 10 seconds).
  2. Add a combined video (regular/static video) to extend the total playback time — this does not consume DVM minutes.

​
Frequently asked questions

We recommend at least 10 seconds. That gives the AI enough audio to generate a natural-sounding voice and keeps the video from sounding abrupt or robotic.
Videos shorter than 10 seconds are prone to quality issues, and videos under 5 seconds often fail completely because there isn’t enough voice input for the AI to work with.
The whole video. If your dynamic video is 12 seconds and the personalized greeting is only 1 second, you’re still charged the full 12 seconds.
Record the dynamic, personalized portion (≥ 10 seconds), then add a combined (regular/static) video to extend the total playback time. The combined video does not consume dynamic video minutes.
Updated July 2026.