Grok Imagine Video on Siray: t2v and i2v for Social Clips

xAI's Grok Imagine video endpoints are now on Siray: x-ai/grok-imagine-video-t2v and x-ai/grok-imagine-video-i2v, both at $0.043 per second in 480p with the current 15% discount. They share a family name with the Grok Imagine image endpoints, not a schema. The video family accepts seven aspect ratios with no auto, and none of the four phone-screen ratios that make the Grok Imagine image aspect-ratio guide useful for wallpapers. A ratio table copied from the image side will fail here.

Key TakeawaysTwo endpoints, x-ai/grok-imagine-video-t2v and x-ai/grok-imagine-video-i2v, both on POST /v1/video/generations.Five fields, all required, none optional: model, prompt, duration, size, aspect_ratio. The i2v variant adds a required image.duration takes any integer from 1 to 15. It is not a menu of preset lengths.size is 480p or 720p, and billing is strictly linear: $0.043 per second at 480p, $0.06 per second at 720p.No seed, no negative_prompt, no audio.

Two Endpoints, Five Fields, Nothing Optional

Text-to-video and image-to-video share a schema and a price. A t2v call needs model, prompt, duration, size, and aspect_ratio. An i2v call needs the same five plus image.

That is the whole request surface: no optional field to skip, no seed to hold a look steady, no negative prompt to steer away from an artifact. Where other video APIs expose a dozen knobs, Grok Imagine exposes five and requires all of them. For social clips that is mostly a benefit: the call is short enough to write by hand.

Duration Is an Integer, Not a Menu

duration accepts any integer from 1 to 15 seconds. Many video models sell fixed lengths — five seconds or ten — which forces a choice between a clip that is too short and a bill that is too high. Grok Imagine bills the length you ask for.

That matters at the short end. A three-second loop or a six-second beat cut to a track is priced as three or six seconds, not rounded up to a preset. The same field covers a full fifteen-second spot, the practical ceiling for most social placements.

Seven Ratios, No auto

The video endpoints accept 1:1, 9:16, 16:9, 4:3, 3:4, 3:2, and 2:3.

Three things are easy to get wrong when the image endpoints are the reference point:

• There is no auto. The image endpoints let the model choose; here you choose.
• The four phone-screen ratios (9:19.5, 19.5:9, 9:20, 20:9) do not exist on the video side.
• The list is shorter overall — seven against thirteen.

For social work the load-bearing entries are 9:16 for vertical feeds and 1:1 for square; 16:9 covers landscape embeds. The other four are compositional conveniences, not distribution formats.

What the Cost Table Actually Says

Billing is per second and strictly linear, with the current 15% discount applied. The Siray console lists it per second rather than per run:

The discount is a promotional price and can be withdrawn, so treat the list column as the planning number. Fifteen seconds at 720p comes to $0.893 today, $1.05 at list. For comparison across the rest of the catalog, see cheap AI video and image APIs.

Since duration bills linearly, the cheapest useful experiment is short: three seconds at 480p is about $0.13, enough to learn whether a prompt describes motion the model understands before paying for fifteen.

A First Call

POST https://api.siray.ai/v1/video/generations
{
  "model": "x-ai/grok-imagine-video-t2v",
  "prompt": "handheld tracking shot, person walking through a night market, neon signage, shallow depth of field",
  "duration": 6,
  "size": "720p",
  "aspect_ratio": "9:16"
}

Six seconds at 720p, vertical, $0.36.

Extending a Clip

x-ai/grok-imagine-video-extension takes a source video plus a prompt and a duration from 2 to 10 seconds, and carries no size or aspect_ratio — it inherits both from the input. It runs at $0.051 per second with the discount. If you need to reach a length the extension route does not cover, Flux 3's video extension is the other path on Siray.

Create your free Siray account and run your first vertical clip on Grok Imagine Video.

A Note on Compliance

Siray's NSFW and uncensored models generate legal, consensual adult content for verified adults only. CSAM and any other illegal content are refused with zero tolerance. Grok Imagine Video is a third-party model from xAI available through Siray's API, not a Siray-hosted model, and the same standard applies to everything generated with it.