Google: Veo 3.1 Lite

Google: Veo 3.1 Lite is a video generation model from Google. It sits in the video category and is described by the provider as Google's most cost-effective video generation model, designed for high-volume applications and rapid iteration. The provider release date is 2026-04-23.

The model accepts text and image inputs and produces video as its only output. It generates video at 720p and 1080p, in 16:9 and 9:16 aspect ratios, with durations of 4, 6 or 8 seconds. It supports first_frame and last_frame image inputs, so you can anchor a clip at its start, its end, or both. Native synchronized audio generation is available. No context length is listed for this model, so token-based context planning does not apply in the usual way; the practical constraints are the resolution, aspect ratio, duration and frame image options above.

Those capabilities point to a fairly specific set of uses. Text to video suits short clips for social feeds, ads and prototypes, and the 9:16 ratio fits vertical formats directly. Image to video with a first frame is useful when you already have a still, such as a product shot or a storyboard panel, and want it to move. Supplying a last frame as well gives you a way to steer a clip toward a defined ending, which helps when several shots need to cut together. Because audio is generated in sync, you can produce a finished short clip without a separate audio pass. The 4, 6 and 8 second durations keep outputs short, which fits iteration: generate several variants, review them, and refine the prompt or the frame images.

You call the model through api14's OpenAI-compatible API using the model id "google/veo-3.1-lite". The supported request parameters are max_tokens, response_format, seed, temperature and top_p. The seed parameter is worth using when you want reproducible or comparable generations across runs, and temperature and top_p give you control over variation between outputs. Because the API is OpenAI-compatible, existing OpenAI client libraries and request shapes work with only the model id and endpoint changed.

There is no listed context length, and maxOutput is 0 in the model details, so do not assume a text output budget. Output is video only, which means tasks that need text, code or still images are a better fit for a model in those categories. If you need resolutions or aspect ratios beyond 720p and 1080p or 16:9 and 9:16, or durations outside 4, 6 and 8 seconds, check whether another video model covers them. For high-volume generation and fast iteration at the listed settings, this model is the intended choice.