minimaxiH3 Get access
xAIText-to-videoVendor content policy applies

Grok Imagine Video Text-to-Video

xAI Grok Imagine Video generates short videos (1-15s) from natural-language prompts at 480p or 720p.

$0.075per secondStarting price at the base resolution and quality tier.
$0.60Typical 8-second clip at the base tier. Longer or higher-resolution outputs scale with duration and tier.
Pay per useBilled per request from your balance. No subscription, no minimum.
Sample output · prompt: “A vast desert megacity at dusk, monumental brutalist architecture stretching endlessly beneath a hazy amber sky, cinematic atmosphere inspired by Denis Villeneuve’s visual storytelling, a lone figure in a flowing dark cl…”

What it does

Grok Imagine Video Text-to-Video, in practice.

Grok Imagine Video Text-to-Video is xAI's text-conditioned video generation endpoint within the broader Grok Imagine multimodal generative system. This README applies to the following API model identifier: Developed by xAI and built atop technology acquired from the Hotshot startup, Grok Imagine is powered by the "Aurora" engine — a unified autoregressive Mixture-of-Experts model that natively interleaves text, image, video, and audio tokens. The text-to-video endpoint converts natural-language prompts into short, audio-synchronized video clips with cinematic motion, ambient sound, music, and dialogue generated in a single forward pass. Within the field, xai/grok-imagine-video/text-to-video

  • Autoregressive Mixture-of-Experts Architecture: Unlike the diffusion transformers used by most competitors, Grok Imagine Video predicts the next token across interleaved streams of text, image, video, and audio. This unified token-prediction design enables a single backbone to serve five conditioning modes (text-to-image, image-edit, text-to-video, image-to-video, video-edit) and dramatically reduces latency relative to iterative denoising pipelines.
  • Native Synchronized Audio Generation: Music, sound effects, ambient noise, and dialogue with lip-sync are generated in the same autoregressive pass as the visual stream, rather than being dubbed in after the fact. This produces tight audio-visual coherence that is difficult to achieve with separate video and audio models.
  • High-Throughput, Low-Latency Inference: Typical generations complete in approximately 17–30 seconds — roughly one-half to one-quarter the time of leading diffusion competitors — making the endpoint practical for interactive ideation and high-volume social-content workflows.
  • Flexible Output Configuration: Supports clip durations up to 10 seconds in consumer products and 15 seconds via API, at 480p or 720p (with 1080p in Pro preview), at 24 fps, across multiple aspect ratios including 16:9 and 9:16. Up to four concurrent video variants can be requested per API call.
  • Long Prompt Support: API requests accept prompts up to 10,000 characters, allowing detailed shot descriptions, camera-motion directives, style references, and dialogue scripts to be included in a single conditioning string.
  • Trained at Frontier Scale: The Aurora backbone was trained on xAI's Colossus supercomputer using a cluster reported at 110,000 NVIDIA GB200 GPUs, enabling the large-scale multimodal token training required for native cross-modal coherence.

Run Grok Imagine Video Text-to-Video

from $0.075/sec
resolution
aspect_ratio
Invite code opens Chat with this model loaded. No code yet? Join the waitlist — we count which models people ask for.

Parameters

What you can set.

ParameterWhat it doesOptions
promptNatural-language description of the video to generate.
durationLength of generated video in seconds. Range: 1–15.default 8
resolutionOutput resolution.480p 720p
aspect_ratioOutput aspect ratio.1:1 16:9 9:16 4:3 3:4 3:2 2:3

Sample prompt

The prompt behind the sample.

A vast desert megacity at dusk, monumental brutalist architecture stretching endlessly beneath a hazy amber sky, cinematic atmosphere inspired by Denis Villeneuve’s visual storytelling, a lone figure in a flowing dark cloak walking slowly through colossal corridors illuminated by soft volumetric light, massive spacecraft hovering silently overhead, drifting dust particles, minimal yet powerful composition, slow cinematic camera movement, ultra realistic textures, deep shadows, atmospheric perspective, muted earth-tone color palette, epic scale, emotional isolation, IMAX framing, subtle lens di
duration: 8resolution: 720paspect_ratio: 16:9

FAQ

Short answers.

How much does Grok Imagine Video Text-to-Video cost?

Pricing starts at $0.075 per second of video at the base resolution; higher resolutions and longer durations cost more. An 8-second clip at the base tier is about $0.60. Usage is billed per request from your balance — no subscription.

Does Grok Imagine Video Text-to-Video run uncensored here?

No. xAI applies its own content policy to this model regardless of where it is called from. For a route without an extra platform filter use Uncensored MiniMax H3; this page lists Grok Imagine Video Text-to-Video for everything else it does well.

What does Grok Imagine Video Text-to-Video take as input?

It is a text-to-video model. xAI Grok Imagine Video generates short videos (1-15s) from natural-language prompts at 480p or 720p.

How do I use it?

Enter an invite code to open Chat with the model loaded, or join the waitlist. We open seats in batches and track which models are requested most.

Related

Models people compare with this one.

Run Grok Imagine Video Text-to-Video.

Invite code opens Chat with the model loaded. No code — join the waitlist and we will count the request.

Get access Browse models