Some aspects of the disclosure provide a method for video generation. The method includes displaying, in response to an audio input operation in an audio input interface provided by a terminal device, a keyword selection interface based on a first audio that is obtained from the audio input operation. The keyword selection interface includes at least a recommended keyword that is obtained by performing a speech recognition on the first audio. The method further includes determining, in response to an editing operation on the at least the recommended keyword in the keyword selection interface, at least a target keyword, and displaying, in response to a video synthesis operation in the keyword selection interface, a video display interface, the video display interface includes a synthetic video from a video clip, and the video clip matches the target keyword.