The Google-research/frame-interpol is a public API tool that allows users to run models to interpolate frames in large scenes of motion, with a minimum output of 1 and maximum of 8 interpolations per video, at a FPS of 30. The tool presents a single-network approach that achieves state-of-the-art results without using additional pre-trained networks such as optical flow depth. The model is trained with a frame triplet alone and is typically completed within 36 seconds using NVIDIA T4 GPU hardware. Additional information, such as citations and examples, are available on the tool’s GitHub page.