Google DeepMind unveiled Veo at Google I/O on 14 May 2024, a text-to-video model for 1080p clips a minute or longer [1]. Veo 2 followed on 16 December 2024, which Google says reaches 4K [2], and Veo 3 at I/O 2025 on 20 May was the first to generate synchronized dialogue, sound effects and ambient noise [3]. Veo 3.1 and 3.1 Fast launched on 15 October 2025 with up to three reference images and scene extension [4]. Veo 3.1 Lite, on 31 March 2026, is the cheapest at $0.05 a second for 720p with audio [5]. Google has not announced a Veo 4 [6].