What is Grok Imagine?

Grok Imagine is an image and video generation model developed by xAI and offered inside the Grok apps, with a public API launched on Jan 28, 2026 and the grok-imagine-video-1.5 model with reference-to-video released on Jul 31, 2026. It is a closed model governed by xAI's acceptable use policy, which permits some adult content but bans minors and non-consensual depictions.

The consumer app markets a less restricted mode for adult subscribers; xAI's legal pages read on Sep 3, 2026 describe the behavior in general terms without naming it.

Which Grok Imagine versions exist?

xAI introduced its Aurora image model in Grok in 2024 and opened Grok Imagine to developers on Jan 28, 2026 with image and video endpoints. The release notes list grok-imagine-video-1.5 on Jul 31, 2026, adding reference-to-video and native 1080p, and a quality parameter for grok-imagine-image-2.0 on Aug 28, 2026.

All versions are closed and cloud-hosted; there are no weights, no Hugging Face repository and no local inference path. Consumer access runs through the Grok apps on iOS, Android and X, with heavier limits unlocked by SuperGrok subscriptions.

Does Grok Imagine allow NSFW output?

More than most peers, within limits. xAI's legal FAQ states that Grok may respond with sexual situations, coarse language or violence when users choose certain features or use suggestive language, and the acceptable use policy places the responsibility for such requests on the user.

The same documents draw hard lines: sexualized depictions of minors are prohibited and reported to the National Center for Missing and Exploited Children, and the policy forbids content that violates other people's rights. The xAI pages we read on Sep 3, 2026 do not publish the exact filter rules of the adult mode.

“For instance, if users choose certain features or choose to input suggestive or coarse language, Grok may respond with some dialogue that may involve coarse language, crude humor, sexual situations, or violence.”

— xAI legal FAQ, accessed Sep 3, 2026

How is Grok Imagine used in tools we review?

No tool in our rankings names Grok Imagine as a backbone, and xAI's terms would not permit reselling non-consensual output. We cover the model because readers ask how a mainstream generator with looser adult settings compares with the services we score under how we rate.

The answer is that the line xAI draws, adult content only with consenting adults and never real people against their will, is the same test we apply. The Ainudez review and the video hub show how the tested services handle that line in practice.

What does the acceptable use policy require?

The xAI Acceptable Use Policy, effective Jan 2, 2025, is incorporated into the consumer and API terms. It forbids illegal content, harassment and rights violations, and xAI states that Grok is not directed at children under 13.

For readers of this site the operative point is that opting into adult output does not license depicting a real person without consent. See digital consent and deepfake for how that duty is defined, and image-to-video for the input mode the API exposes.

Sources

1. xAI — Grok Imagine API announcement — Jan 28, 2026, accessed Sep 3, 2026.
2. xAI — Grok Imagine Video 1.5 with references — Jul 31, 2026; reference-to-video, accessed Sep 3, 2026.
3. xAI developer docs — release notes — model version dates, accessed Sep 3, 2026.
4. xAI — Acceptable Use Policy — effective Jan 2, 2025, accessed Sep 3, 2026.
5. xAI — legal FAQ — statement on sexual situations and age, accessed Sep 3, 2026.

Reference entry by the Clothoff AI Editorial Team; not legal advice. The access date marks the last verification.