Kling: Kling O1 Omni Video API

Kling Video O1 (Omni-Video) is a next-generation unified multimodal foundation model designed for professional-grade video synthesis and real-time semantic editing. Built on the "Omni-Temporal Transformer" architecture, it treats video not as a series of frames, but as a continuous space-time latent flow. The O1 model introduces "Conversational Video Editing," allowing users to modify specific elements (e.g., "Make the character smile more," "Change the car to a vintage red sedan") through natural language dialogue while maintaining perfect temporal consistency. It is the first model to achieve "Human-Level Physical Intuition," accurately rendering complex fluid dynamics, cloth physics, and intricate muscle movements in 2026.

• Input: text, image, video
• Output: video
• File input: Supported
• Released: 2025-11-30
• Knowledge cutoff: 2026-02-28