Mistral NSFW and Le Chat NSFW Rules Inside Vibe and the Mistral API

Does Mistral allow NSFW content?

Not in so many words, and that is the interesting part. Mistral's Usage Policy, effective June 11, 2026, has three rules that touch sexual content. It bans child sexual abuse material with "zero tolerance", reports it to law enforcement and terminates the account. It bans generating "intimate images of any person without the explicit consent of all individuals involved". And under illegal activities it bans content that "enables or exploits individuals, including through human trafficking or sexual services".

That is the whole list. We searched the Usage Policy, the US consumer terms and the commercial terms for erotic, pornography, NSFW, sexually explicit and adult content, and none of those words appears. Compare that with Anthropic, which bans "erotic chats" by name, or DeepSeek, which bans "sexual chatbots" by name. On paper, Mistral is the most permissive of the big assistant companies.

On paper is the operative phrase. Silence in a policy is not a permission you can point to later. The same terms let Mistral "review, remove, restrict access to, or disable Your Data at any time" if it "creates an actual risk or liability for us or others", which is broad enough to cover anything Mistral decides it no longer wants to host.

Does Le Chat allow NSFW now that it is called Vibe?

Le Chat is now Vibe. The US App Store listing is titled "Vibe by Mistral (ex-Le Chat)" and opens with "Le Chat is now Vibe", and Mistral's pricing page refers to "Vibe (or Le Chat)". The rules above apply to Vibe exactly as they applied to Le Chat.

What Vibe does not have is any adult setting. We found no adult mode, no age verification step and no content preference in Mistral's documentation. The consumer terms set the minimum age at 13, with "parental or legal guardian permission, where required", and the App Store rates the app 13+. A product that admits 13 year olds is not designed around explicit adult chat, whatever its policy leaves unsaid, and that is the practical reason Vibe answers some adult requests and declines others.

Custom instructions do not change the picture. Mistral documents them as "persistent preferences that shape how Work responds", with no word on content limits, and the terms separately ban "circumventing, or bypassing security or moderation mechanisms". Writing instructions designed to get around moderation is exactly what that clause describes.

Coming from ChatGPT? Its rules are on our ChatGPT NSFW and adult mode page. For DeepSeek, see DeepSeek NSFW, for Microsoft, whose terms reach your whole Microsoft account, Copilot NSFW, and for Meta, which bans erotic chat by name, Meta AI NSFW.

Does the Mistral API allow NSFW roleplay?

This is what most Mistral NSFW searches are really about, because Mistral models sit behind a lot of roleplay frontends. Three facts from Mistral's own documentation decide it.

First, the API is not filtered by default. The chat completions endpoint has a safe_prompt parameter whose default value is false, and Mistral now marks it deprecated. When it was switched on, it prepended a system prompt that told the model to "Avoid harmful, unethical, prejudiced, or negative content." Off by default means a raw API call reaches the model with no injected safety instruction.

Second, moderation is something the developer adds. Mistral sells a moderation model, mistral-moderation-2603, with eleven categories. Its Sexual category covers "Material that explicitly depicts, describes, or promotes sexual activities, nudity, or sexual services", and "includes pornographic content, graphic descriptions of sexual acts". Its Jailbreaking category covers attempts to bypass safety guidelines "through prompt manipulation, role-playing scenarios, or other techniques". Custom guardrails check input only and return a 403 when they trigger. So whether a roleplay app built on Mistral blocks you depends on the app, not on Mistral.

Third, the Usage Policy still applies to the API, including the ban on trying to "circumvent security protections and AI safety filters", and the consumer terms describe the APIs as "limited to business customers". If you are paying for an API key to run personal roleplay through a frontend, you are using a business product under business terms.

If you route roleplay through a frontend, the same question for the frontend side is covered in does Janitor AI need a proxy.

Can you get banned from Mistral for NSFW?

For consensual adult text, Mistral has written no rule you would be banned under. For the categories it does name, the consequences are written down. The Usage Policy says violations "may result in temporary suspension or permanent termination of your account", plus reporting to authorities where appropriate. The consumer terms add that Mistral may "immediately suspend or terminate" an account for breaching its policies or if continued use "could cause risk or harm to Mistral AI, our users, or anyone else".

Two clauses catch people who experiment. The terms ban "circumventing, or bypassing security or moderation mechanisms", so a jailbreak prompt is a terms violation even when the content it unlocks is not. And "the creation or use of multiple Mistral AI accounts by a single individual is strictly prohibited", which rules out a second account kept for adult use. Mistral does monitor: it reserves "the right to monitor your use of Mistral AI Products through automated means". Wrongful suspensions can be appealed at support@mistral.ai.

Is Mistral private enough for adult conversations?

By default, your Vibe chats help train Mistral's models. Mistral's help center says "Input and output data are used by default to train our artificial intelligence models, unless you opt out of training at any time", and that Vibe users "are not opted out by default". Only Enterprise customers start opted out. We found no difference between Free and Pro on this.

You can switch it off. On the web it is the toggle "Allow your interactions to be used to train our models" in the admin panel under Vibe; on mobile it is "Enable data sharing" under Data & Account Controls. Two catches: the Vibe and API opt-outs are separate toggles, and pressing thumbs up or thumbs down with a comment authorizes Mistral to use that conversation whatever your plan. Chats are kept until you delete them or the account, and account data for a year after deletion.

We compared retention and training defaults across the big assistants in how long AI companies keep your chats.

How much is Mistral Pro, and what do you get for adult use?

On mistral.ai/pricing in US dollars, Vibe Free costs nothing, Pro is $14.99 a month before tax, verified students pay $5.99 for up to 12 months if they are new to Vibe, and Team is $24.99 per user. The US App Store lists two "Vibe Pro" purchases, $14.99 and $149.99, without saying which period the second covers. Payments are "non-refundable except where required by applicable law", subscriptions renew automatically, and Mistral promises 30 days notice of a price rise.

Pro buys more messages (up to six times Free), more web searches, up to 40 times the image generations, coding sessions and $30 a month in API credits. None of it is an adult feature. Paying does not change the age floor, the moderation model or the policy, so for adult chat Pro and Free are the same product with different limits.

For the whole market's prices in one place, adult AI chat pricing compares them.

What Mistral does better than we do

A lot, and you should know it before paying anyone. Vibe is a full assistant: it codes, searches the web, reads your files, connects to more than 100 apps and generates images. Its API is cheap, with Mistral Small 4 at 15 cents per million input tokens. And Mistral publishes open models under Apache 2.0 that you can run on your own computer, where its Usage Policy states plainly that it "does not apply". If you have the hardware and enjoy setting things up, that is the most private and least restricted way to use a Mistral model, and nothing we sell beats it on freedom.

We do one thing: private adult conversation with a companion who remembers you, open only to adults, with explicit content allowed in our own rules. If you want both, keep Mistral for work and use us for the rest.

How Depravity works as a Mistral alternative for adults

You open a chat and talk. There is no API key, no guardrail setting, no training toggle to hunt for and no token meter, and every plan is unlimited chat.

• Adult from the first message. Accounts are 18 and over and our rules allow explicit content between consenting adults, so there is no filter to negotiate with.
• Written limits you can read. Nothing involving minors, nothing non-consensual, nothing illegal. That is the complete list.
• Private by default. Chats are encrypted, never sold, never shared and never used to train advertising, and you can export or delete them. We use automated and human review to enforce the three hard lines, and we say so.
• A companion, not an assistant. The character keeps its voice and remembers names and history across sessions.

Plans and yearly prices are on pricing. For a ranked comparison of the adult options, see best uncensored AI chatbots.

Which option should you pick?

Match it to what you actually want. We would rather point you elsewhere than sell you the wrong thing.

• You want a strong assistant for work, coding or research: Vibe Pro at $14.99 is good value, and nothing here competes with it.
• You have a capable GPU and like tinkering: run Mistral's open weights locally, where its service rules do not apply.
• You want long adult fiction written as prose with a manuscript view: a dedicated writing tool will serve you better than a chat.
• You want adult conversation or roleplay with a character, adults only, explicit content allowed in writing and no account at risk: Depravity, from $12 a month billed yearly.

Writing prose rather than chatting? Our AI smut writer page compares the tools built for manuscripts. Comparing Google's rules instead? See Gemini NSFW.