Kling: kling_avatar_image2video API
Developed by Kling AI, this model specializes in generating high-fidelity talking avatar videos. By taking a single portrait (human or stylized) and an audio track as input, it produces videos with precise lip-sync, natural facial expressions, and subtle upper-body movements. The price is 1 second for standard (std) mode, double the price for pro mode.
- Input: audio, image, text
- Output: video
- File input: Supported
- Released: 2025-10-14