Grok Imagine

True text-to-video (earlier Grok video was image-to-video only), reference inputs, native 1080p; a 6s 720p clip in ~25s, down from 40s+; $0.08 per second

License

Proprietary

Commercial use

Allowed

Availability

Official API, Consumer app

Audio

sound effects, ambience and dialogue generated in the same pass

Resolution

native 1080p

Max duration

15s

Preview 2026-05-31, generally available 2026-06-16. Dated to the preview announcement, matching how Wan 3.0 is dated to its public beta.

Higher realism tier for the Grok Imagine API; stronger text rendering; tighter prompt following; improved world understanding; enterprise-focused (product shots, UGC-style content, brand assets)

License

Proprietary

Commercial use

Allowed

Availability

Official API

Model ID

grok-imagine-image-quality

Enterprise API tier launch. Pairs with grok-imagine-video for combined image+video workflows.

Unified image + video generation, editing, and restyle in one API; virtual try-on; product placement; image and video editing from text prompts; topped Artificial Analysis T2V leaderboard at launch

License

Proprietary

Commercial use

Allowed

Availability

Official API, Consumer app

Model Id Video

grok-imagine-video

Resolution Image

2K (up to 10 images/request)

Max Duration Video

15s

Described on launch as 'state-of-the-art video generation, video editing, and image creation, all in one API'. xAI joined SpaceX Feb 2, 2026.

First xAI image generation; codename Aurora; native multimodal input (image editing from input images); available on X platform; strong photorealism and text following

License

Proprietary

Commercial use

Allowed

Availability

Consumer app

Architecture

Autoregressive Mixture-of-Experts (trained on interleaved text and image data)

Released on X platform December 9, 2024. Superseded by Grok Imagine API (Jan 2026).