Kling: Kling v1.6 Multi Image to Video API

Kling v1.6 Multi-Image-to-Video is a specialized variant of the v1.6 family that leverages the "Elements Library" to achieve superior subject consistency. Unlike standard I2V which uses a single start frame, this model allows creators to upload multiple reference images (such as different angles of a character, clothing details, or specific objects). The system extracts deep semantic features from all provided images to ensure the generated motion adheres to the character's identity with a 195% improvement in consistency compared to v1.5. In 2026, it is widely used for professional "Virtual Try-On" and "Character Re-animation" where maintaining the visual integrity of a brand or persona is critical.

• Input: text, image
• Output: video
• Released: 2024-12-18
• Knowledge cutoff: 2026-02-28