Rogue Studio ships an uncensored video model and prices permission from $29 to $299

Rogue Studio, a cinematic AI generation platform built around adult visual storytelling, launched the day WIRED published its account of it, with a marquee video tool called Rogue 1.0 that WIRED describes as comparable to the pro-level models Hollywood is already using ^[1][2]. The interesting part is not the pixels. It is the product thesis: the constraint being sold here is not capability but permission, and it now has a price list, from $29 to $299 a month ^[3].

That spread is roughly a tenfold gap between entry and top tier ^[17], which suggests the company expects volume users rather than curious ones. The bundle is telling too. Rogue carries safe-for-work frontier models inside its own system, including NanoBanana, Seedance and Kling 2.0, in an arrangement WIRED likens to Higgsfield, alongside proprietary models for adult images and video ^[4]. WIRED's example workflow has a user ideating a character in Midjourney and then using Rogue's tools to add uncensored high-definition detail ^[5]. In other words, the general-purpose models do the work everyone can already do, and the paid differentiator is the step the others refuse.

The cofounder who goes by Mr. Rogue frames the gap in market terms, telling WIRED that a sophisticated creator wanting R-rated work has a choice between "skeezy NSFW sites that are built on outdated open-source tech" and high-quality tools that are "censored up the gills," and that the ambition is "what HBO did for television" ^[6][7]. The studio also bills itself as a "playground for creative ethical mischief" ^[8].

The compliance apparatus is where the claim gets tested. Rogue's terms prohibit using customer images or videos for training and forbid nonconsensual deepfakes, child sexual abuse material, unauthorized depictions of public figures and other illegal or abusive imagery ^[9]. Enforcement, per WIRED, is three layers of human moderation: keyword flagging at the prompt level that can trigger account suspension, an "intention layer" where moderators reread prompts for tricks such as misspelling a real actor's name, and a final review of the generated output, so that a prompt for a "famous blonde singer" that returns something resembling Taylor Swift is not delivered ^[10]. Human review at three checkpoints is a cost line that grows with usage, which is the thing to model before you believe the tiering.

Both cofounders are movie producers who declined to be named, citing "real risk" in being associated with erotic work and the current climate around Hollywood and AI; WIRED says it verified that they work in Hollywood on films with combined box office over $2 billion ^[11]. So the operators are commercial and specific about their product while remaining pseudonymous about themselves, which is its own signal about where reputational risk currently sits.

Context supports the demand case without confirming it. WIRED reports demand for AI-generated adult material is surging while quality on companion apps is not movie-caliber ^[12]. Netflix has said 300 of its shows use generative AI ^[13], and Puck founding partner Matt Belloni told WIRED that integration is happening at a larger scale than viewers realize ^[14]. A 2023 study by The Economist, cited by WIRED, found sex scenes in movies down 40 percent since 2000 ^[15]. Post-strike, unions won guardrails on unapproved digital likenesses while studios examined AI as a cost lever ^[16].

Watch whether the third-party model providers inside Rogue's stack stay comfortable sitting next to an uncensored tier, whether the human moderation layers survive contact with paid volume, and whether any operator in this segment puts real names on the door.