Video generation is the one place where you could not choose anything. Which model produced your video, and how many videos a single click produced, were both decided for you — and the number was a consequence of how many models happened to be switched on rather than a setting anyone had chosen. In practice every click produced two videos from two different models and charged for both.
Team settings now hold two new controls, next to the default image model and the packshots count that were already there. Default video model selects the model every video generation will use, with its credit cost shown beside each option. Videos per generation sets how many videos one click produces, from one to four. All of them are submitted together and appear in your list at the same time; how many run at once still depends on your plan's limits.
Both settings are edited by an account owner and apply to everyone on the team, including generations started through the plugin.
What changes if you do nothing. A team that never opens these settings now gets one video per generation instead of two, at the price of one video instead of two. Fewer outputs and a lower cost, with no action required — and if you preferred two, that is one click in settings.
Video models are also named the way image models already were. You will see names like Qamera Aurora and Qamera Nimbus in the selector, on the tiles and in the generator filter, in place of the engine names that were showing through before. The names carry no numbering and no version, because these models are alternatives rather than a ranking — the credit cost next to each one is what tells them apart. Videos you generated earlier keep working and are relabelled along with everything else.
One more fix in the same release: the video and packshot creation windows now close once generation starts. They used to stay open with your image still selected, which made it easy to click again and pay for a second batch by accident.