Acceptable Use Policy

Last updated: 2026-08-01

This Acceptable Use Policy ("Policy") governs what you may and may not do with MiniMax H3 (the "Service"). It supplements our Terms of Service. By using the Service you agree to this Policy. We may suspend or terminate access, refuse or remove any generation, and take other enforcement actions for conduct we believe violates it.

1. Independent AI interface

MiniMax H3 is an independent product that provides access to third-party AI models through our own interface. We are not affiliated with, endorsed by, or sponsored by the providers of those models. Model and company names are used for identification only.

2. Prohibited content categories

You may not use the Service to generate, upload, request, or distribute any of the following categories of content (listed explicitly for clarity):

1. Sexual / NSFW content — sexually explicit, pornographic, or sexually suggestive material, whether or not subjects appear to be adults.
2. Violence / gore — graphic violence or gore for shock value, or content that incites real-world violence.
3. Hate speech — content that promotes hatred or violence against people based on a protected characteristic.
4. Child sexual exploitation (CSAM) — any sexual content involving minors (anyone 17 or under), including fictional depictions. This is zero-tolerance and will be reported to authorities where required.
5. Deepfakes / impersonation — non-consensual depictions of real, identifiable people, including face-swap or other face-manipulation of real people, or misleading deepfakes.
6. Copyright / trademark infringement — content that infringes intellectual property, publicity, or privacy rights.

2.1 Zero-tolerance (immediate action)

We treat the following as grounds for immediate suspension or ban, and may report to law enforcement:

• CSAM or any sexual content involving minors
• Content that promotes or assists terrorism, violent extremism, or mass violence
• Guidance for producing weapons of mass destruction
• Non-consensual intimate deepfakes of real people

2.2 Other prohibited conduct

You may not:

• generate or spread scams, malware, phishing, or other deceptive practices
• misrepresent AI-generated output as human-created where that is deceptive or unlawful
• impersonate real people, brands, or institutions to mislead others
• use jailbreak prompts or other techniques to bypass safety or moderation systems
• resell or expose the Service as an “uncensored”, “unfiltered”, or “no-filter” generation tool
• use automated means to overload, scrape, or abuse the Service beyond your plan
• bulk-generate spam or fraudulent content

3. Content moderation

We use a layered approach appropriate to an AI generation product:

3.1 Automated checks

• Input screening: prompts and relevant inputs may be scanned before generation (keyword, classifier, and/or third-party moderation APIs).
• Output screening: generated results may be evaluated before or after delivery.
• Provider safety filters: upstream AI model providers may also block or refuse requests under their own policies.
• Usage signals: abnormal traffic or repeated policy violations may be rate-limited or blocked.

Blocked requests may return an error and typically do not consume credits when moderation rejects the request before generation.

3.2 Human review

Flagged items, user reports, and serious safety cases may be reviewed by staff. We may preserve moderation logs and related records for security, abuse prevention, and legal compliance (commonly on a rolling retention window).

3.3 Limits of automation

No system catches every violation. You remain responsible for your prompts and how you use outputs. We may take action on content we discover later even if it initially passed automated checks.

4. Reporting

4.1 What to report

• content or behavior that may violate this Policy
• CSAM, terrorism, or other severe abuse
• AI impersonation or infringement of your rights

4.2 How to report

Email support@minimax-h3.net with:

1. the URL or description of the content / account behavior
2. why you believe it violates this Policy
3. any evidence you can share

We treat reports confidentially to the extent practicable, review them in good faith, and may record abusive or false reports.

5. Enforcement

Depending on severity and history, we may take one or more of the following actions:

We may also cancel unused credits or refuse refunds where termination is for Policy violations, as allowed by the Terms and Refund Policy.

6. Appeals

If you believe an enforcement action was made in error, email support@minimax-h3.net within 14 days of notice with your account email and a brief explanation. We will re-review when feasible. Appeals for zero-tolerance illegal content may be limited.

7. Changes

We may update this Policy from time to time. Continued use of the Service after a change takes effect constitutes acceptance of the updated Policy.

Questions? Contact support@minimax-h3.net.