Grok “Spicy” mode is an opt-in content setting inside Grok Imagine, the image and video generator built into xAI’s Grok assistant. When selected, it loosens the tool’s default content filters to allow more suggestive, mature-leaning output than the standard settings permit, while still stopping short of fully explicit material and remaining locked behind a paid subscription and an 18+ age check.
What Grok Spicy mode actually does
Grok Imagine is xAI’s generative tool for turning text prompts and still images into pictures and short video clips. Like most mainstream image generators, it applies content moderation by default so that outputs stay within a general-audience range. Spicy mode changes that ceiling. Selecting it tells the system to permit “bolder” results, which xAI has framed as closer to R-rated film content than to hardcore pornography.
In practice, that means Spicy mode can produce suggestive poses, partial nudity, and more intimate framing than the standard modes allow. xAI’s stated policy still blocks certain categories outright, including fully explicit sexual acts, any sexual content involving minors, and child sexual abuse material. The company has also publicly said it prohibits sexualized depictions of real, identifiable people without consent, though, as covered below, enforcement of that line has been a major point of contention.
Spicy is not a separate app or a hidden jailbreak. It is a labeled option that appears in the generation interface once your account meets the eligibility requirements. If your account does not qualify, or you are in a region where the feature is restricted, the option simply does not show up.
The four Grok Imagine modes
When you create or animate an image in Grok Imagine, the tool typically offers a small set of content modes. Each one shifts how tightly the output is filtered and what stylistic direction the model leans toward.
| Mode | What it does |
|---|---|
| Normal | The default, general-audience setting. Produces straightforward images and video with standard moderation applied. |
| Fun | A more playful, exaggerated interpretation of a prompt, adding motion or whimsical flourishes while staying within general-audience limits. |
| Custom | Lets you steer the result with your own text instructions for style, motion, or tone, rather than relying on a preset. |
| Spicy | Relaxes default filters to allow suggestive, mature-leaning output such as partial nudity. Gated behind a paid plan and 18+ age verification. |
The exact wording and lineup can shift as xAI updates the app, but Normal, Fun, Custom, and Spicy have been the recognizable set since the feature rolled out. Spicy is the only one of the four with hard eligibility gates attached.
How to turn on Grok Spicy mode
Because the setting is age-restricted and paywalled, enabling it involves a few distinct steps rather than a single toggle. The general process looks like this:
- Get the app and a qualifying subscription. Spicy mode has generally required a paid tier, either SuperGrok or X Premium+. The lower-cost base X Premium plan typically does not unlock it. Pricing and plan names change over time, so confirm current terms in the app before subscribing.
- Use the Grok mobile app. The feature has primarily been available in the Grok app on iOS and Android. Web and desktop access has been limited or removed in many regions, so mobile is the reliable path.
- Verify that you are 18 or older. xAI requires age verification to access Spicy content. Depending on your country, this can range from a date-of-birth confirmation to submitting an ID document, particularly where local law mandates stricter age checks.
- Enable sensitive-content preferences. In your account or content settings, turn on the options that allow sensitive media to be displayed and generated. These act as a master switch that must be on before the mode appears.
- Open Grok Imagine and select Spicy. Once the above conditions are met, generate or animate an image and choose Spicy from the mode selector. If it is missing, your plan, age status, or region likely does not currently qualify.
One important caveat: availability has been inconsistent. Users in various countries have reported the option appearing and disappearing as xAI adjusts its rollout and responds to regulatory pressure. It is not guaranteed to be present everywhere, even for eligible accounts.
Regional restrictions
Where Spicy mode works depends heavily on local law. xAI gates the feature by region, and access has been narrowed or blocked entirely in several markets. Some countries in Southeast Asia moved to restrict access, and regulators in India pressed X over the broader issue of AI-generated obscene and non-consensual content in late 2025 and early 2026. In the United Kingdom and parts of the European Union, stricter age-verification and online-safety rules shape how, or whether, the feature is offered. The practical takeaway is that eligibility is not just about your subscription and age, it is also about where you are.
Controversy and safety
Spicy mode has been one of the most heavily criticized AI features of the past year, and the concerns are substantive rather than merely reputational.
The most prominent flashpoint came shortly after launch, when reporters and users found that the tool could generate sexualized deepfake imagery of real, identifiable celebrities, including widely reported examples involving Taylor Swift. That directly cut against xAI’s stated policy barring non-consensual sexual depictions of real people, and it fueled accusations that the company’s safeguards were inadequate or inconsistently enforced. Critics argued that a feature capable of producing such material at scale, with weak guardrails, was foreseeable rather than accidental.
The deeper problem observers have highlighted is non-consensual intimate imagery, or NCII. Safety researchers and policy analysts have noted that the core harm is the creation of a sexualized image of a person who never consented to it, regardless of whether it is shared. When image generation of this kind is tied to a mass-audience platform like X, the potential for abuse, including against private individuals and minors, grows significantly.
That has drawn regulatory and legal attention. In the United States, non-consensual intimate deepfakes fall under the federal Take It Down Act, and a coalition of state attorneys general has pressed xAI over its handling of such content. India’s technology ministry issued formal demands to X over AI-generated obscene material, and other governments have weighed intervention. Analysts have pointed out that many of these responses lean on existing intermediary-liability and obscenity frameworks rather than purpose-built AI rules, exposing gaps in how quickly regulation can adapt.
xAI, for its part, has maintained that Spicy mode is intended for consenting adults, that it blocks the most harmful categories, and that it uses age verification and content filters to keep the feature within bounds. The company has adjusted availability and moderation in response to pressure. Whether those measures are sufficient remains an open and actively contested question.
The bottom line
Grok Spicy mode is a real, deliberately built feature, not a loophole: an age-gated, subscription-only setting in Grok Imagine that permits more suggestive content than the default modes. Turning it on requires the Grok mobile app, a qualifying paid plan, verified 18+ status, and a region where the feature is offered. But the same permissiveness that defines it has made it a lightning rod. The documented failures around celebrity deepfakes and non-consensual imagery, and the resulting scrutiny from lawmakers and regulators worldwide, mean that anyone evaluating the feature should weigh not just how to enable it, but the serious ethical and legal issues that continue to surround it.
