According to Beating AI news, Google has released Gemini Omni 1.1 Flash, a new version of Gemini Omni Flash that debuted in May and opened for API public testing at the end of June. The new version still allows for a maximum of 10 seconds of video generation at a time, but for the first time, it includes video continuation in the Omni API. After a video is generated, it can now continue filming. Each continuation can add another 10 seconds, allowing for a maximum total video length of 40 seconds after continuous writing. Each time it continues, the model will reference up to 10 seconds of the previous footage to ensure that characters, actions, and plots are as coherent as possible. Omni 1.1 also introduces head and tail frame control. Users can provide a starting image and an ending image, allowing the model to automatically generate the continuous video in between. For example, if the specified shot starts with a character standing in the distance and ends with them coming closer to the camera, the model will complete the entire motion process on its own. Additionally, a cheaper 360p draft mode has been introduced, with the official claim that system throughput has increased by up to 60%, and costs about one-third of 720p. The API pricing is set at $0.03 per second for 360p, $0.10 for 720p, $0.15 for 1080p, and $0.30 for 4K. Both 1080p and 4K are upscaled in post-production and are not generated in native high resolution.