Skip to main content
Lip cloning is Sendspark’s latest lip sync model for AI-personalized videos. Turn it on inside a dynamic video campaign by opening the Settings gear and selecting “Voice + Lip Model (Beta),” then create your videos. For the best sync, record with good lighting, start with a closed mouth, and keep the first 10 seconds meaningful.

​
Key takeaways

  • Enable it in a dynamic video campaign: click the Settings gear in the top right corner and select “Voice + Lip Model (Beta).”
  • The AI prioritizes the first 10 seconds for lip sync calibration, so trim silent or unnecessary opening seconds and keep that section meaningful.
  • Start recording with your mouth closed, and pause and close your mouth again after your greeting to set the rhythm for smoother sync.
  • Good lighting matters — keep your face well-lit and your lips in focus, and follow voice cloning best practices (speak normally, no background noise, pause after saying “watermelon”).
  • Minor lip sync imperfections may appear depending on the content, but most viewers find them subtle, and the Sendspark team keeps improving the model.

​
How to use the latest lip sync model

To use the latest lip sync model for AI-personalized videos, follow these steps:
  1. When creating a dynamic video campaign, select the “Settings” gear in the top right corner
  2. Select “Voice + Lip Model (Beta)” Lip Sync ​​Lip Sync​
  3. Now, you can go ahead and create videos from this campaign, and you will be using the latest model!
How to Get Best Results for Lip Cloning

​
Pre-Recording Preparations

  1. Trim Silent or Unnecessary Initial Seconds: Before submitting your video for processing, ensure the first 10 seconds are meaningful and relevant, as Sendspark’s AI prioritizes this section for lip sync calibration.
  2. Start With a Closed Mouth: Begin recording with your mouth in a neutral, closed position before speaking. This provides the AI with a clear starting point for synchronizing movements.
Lighting is super important!
  • After your greeting or introductory statement, briefly pause and close your mouth again before continuing with the main content of your message. This helps set the rhythm for smoother lip sync throughout the recording.
Make sure your face is well-lit, and your lips are in focus when recording your original video. Also, follow the standard best practices to get the best results from voice cloning (speak normally, make sure there is no background noise, leave a brief pause after saying “watermelon” before the rest of your message). Current Limitations of Lip Cloning: Sendspark’s lip sync feature is a powerful tool. However, minor imperfections in lip movements or lip sync might appear depending on the video content, though most viewers find these subtle and unnoticeable unless closely scrutinized. The Sendspark team is continually working on updates to enhance this technology.

​
Frequently asked questions

When creating a dynamic video campaign, click the Settings gear in the top right corner and select “Voice + Lip Model (Beta).” Videos created from that campaign will then use the latest model.
Sendspark’s AI prioritizes the first 10 seconds for lip sync calibration, so make sure that section is meaningful and relevant. Trim any silent or unnecessary opening seconds before submitting your video for processing.
Start recording with your mouth in a neutral, closed position, and pause and close your mouth again after your greeting before continuing. Keep your face well-lit and your lips in focus, and follow standard voice cloning best practices — speak normally, avoid background noise, and leave a brief pause after saying “watermelon.”
Not always. Minor imperfections in lip movements or sync can appear depending on the video content, though most viewers find them subtle and unnoticeable unless closely scrutinized. The Sendspark team is continually working to enhance the technology.
Updated July 2026.