What is Grok Imagine Video 1.5 Lite, and what does Sume run?

Grok Imagine Video 1.5 Lite is xAI's lower-cost video generation model, listed at $0.02 per second of output on xAI's models page. It takes a text prompt or a starting image and can speak lip-synced dialogue. xAI says it reaches 1080p by upscaling. Sume does not list a Lite row: the Sume video catalog has grok-imagine-video-1.5, which accepts a source image only.

What does xAI say Lite is?

xAI's video generation page names three models: grok-imagine-video-1.5, grok-imagine-video-1.5-lite and the older grok-imagine-video. It describes the 1.5 model as native 1080p and the Lite model as the cheaper option that reaches 1080p through upscaling. Clip length runs from 1 to 15 seconds with a default of 8, and audio is generated by default on both 1.5 models, with a generate_audio switch to turn it off.

xAI's own recommendation on that page is to use Lite for simple videos, because it has the lowest price per second.

• Modes on the video generation page: text-to-video, image-to-video, edit-video and extend-video. Reference images, first and last frame and keyframes are marked as 1.5-only.
• Resolutions: 480p (the default), 720p and 1080p.
• Aspect ratios listed: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9 and 5:2.

Is Lite on Sume?

No. Sume's video router documentation lists grok-imagine-video-1.5 with 480p and 720p, durations of 4 to 15 seconds, and a note that it is image-to-video only: the frame follows the aspect ratio of the source image you send. There is no text-only Grok request and no Lite id. If you send a model id that is not in the catalog, the generate route rejects it, so check the live catalog before you wire a pipeline around a model name you saw in the news.

Sume's catalog rows are listed at the provider price times 1.25, and the Grok row bills a flat 1.25 cents per output second.

How do the two compare on a 10-second clip?

The table puts the published xAI per-second prices next to the Sume catalog rate. They are different models, so this is a budget comparison and not a quality claim.

Which should you pick?

If you need text-to-video from a bare prompt, Sume's Grok row cannot do it; pick a Sume model that accepts a prompt alone, or generate a still first and animate it. If you already have a product photo or key art, the image-to-video Grok row is the cheapest way on Sume to turn it into a 4 to 15 second silent clip. Submit it through the Videos endpoint, which returns a job id you poll until it completes.