[DigitalToday reporter Yoonseo Lee] Google has unveiled Veo 3.1 Lite, a low-cost API for its Veo 3.1 video-generation artificial intelligence (AI) model.

On April 2 local time, online media outlet Gigazine reported that Veo 3.1 Lite supports text-to-video (T2V) and image-to-video (I2V) generation. Users can create videos with audio based on a text prompt or an image. Supported resolutions are 720p and 1080p. For 720p, video length can be selected as 4 seconds, 6 seconds or 8 seconds, while 1080p is fixed at 8 seconds.

It appears to be designed to let users use video-generation capabilities at a lower cost by reducing features compared with higher-end models. Because it supports both text and image methods, it is interpreted as a model targeting demand for relatively simple video production or prototype work.

But the model excludes 4K video generation, video reference-based generation (V2V) and video extension features supported by the higher-end Veo 3.1 and Veo 3.1 Fast. Advanced functions such as high-resolution output, editing based on existing videos and generating long sequences are not included in the supported range.

As a low-cost model, it offers price competitiveness. API fees are set at $0.05 per second for 720p (about 75 won) and $0.08 per second for 1080p (about 120 won). As a result, it could be seen as an option that lowers barriers to entry for developers or service operators that value cost efficiency over advanced features.

By offering the Veo 3.1 series in higher-end and low-cost versions, Google appears to be pursuing a strategy of running performance-focused models and cost-efficient models in parallel in the video-generation AI market. Detailed features and how to use each model can be found in the Gemini API documentation.