Pika H3 Flex
Pika H3 Flex Reference to Video: generate a 5 to 15 second video with native audio guided by image, video and audio references, at 480p or 768p, or upscaled to 1080p, 2K or 4K. Billed per output second, plus reference video seconds.
pika/pika-h3-flex/reference-to-video