Acceptable Use Policy
Last updated: 24 April 2026
Orpheus is built for legitimate creative, educational, and commercial narration. A few things it is not built for. Using the Service in any of the ways listed below is a material breach of our Terms of Service and can result in immediate account termination without refund.
1. Identity and consent
- Do not generate audio impersonating a real identifiable person (politician, celebrity, acquaintance, yourself-to-defraud-someone, etc.) without their express, informed, written consent.
- Do not generate audio designed to deceive a listener about who is speaking (voice-phishing, fake statements attributed to a real speaker, deepfake news).
- Do not use Orpheus in automated voice-scam / vishing / "grandparent scam" operations.
1a. Voice cloning (Pro tier)
- You may only upload reference recordings of voices you have a legal right to clone — your own voice, or a voice whose owner has given you express, informed, written consent for AI synthesis use.
- You may not upload a recording of a public figure, celebrity, customer, employee, family member, or any other identifiable person without that person's explicit consent for voice cloning.
- Audio generated from a clone is subject to all the rules in Section 1 — consent doesn't authorise impersonation, fraud, or deceptive content.
- If we receive a credible takedown notice from a person whose voice was cloned without consent, we will delete the clone, the generated audio, and (if abuse is established) terminate the account responsible.
2. Illegal content
- No content that is illegal under the law of Greece, the European Union, or your country of residence — including but not limited to: child sexual abuse material (CSAM), incitement to terrorism, content that threatens or glorifies violence against identifiable persons or groups, human trafficking, unlawful stalking or harassment.
- No defamatory content targeting identifiable private individuals.
- No content designed to facilitate the unauthorized distribution of controlled substances, weapons, or CSAM.
3. Intellectual property
- Do not submit text you do not have the right to submit — copyrighted books, articles, or lyrics without a licence from the rights holder.
- Audio you generate from text you legitimately own, or from text that is in the public domain, is yours to use commercially.
4. Technical abuse
- Do not attempt to circumvent rate limits, the fair-use cap, or authentication. Automated tools are welcome within the published limits.
- Do not share API credentials across multiple organisations or resell raw API access.
- Do not attempt to reverse-engineer voice models, extract model weights, or crawl the Service to mass-clone outputs.
- Do not use the Service as a transport for malware, spam, or phishing payloads.
5. Commercial redistribution
You may use generated audio in commercial products (videos, podcasts, audiobooks, training material, client work). You may not re-sell Orpheus itself, re-package the Service as your own, or build a directly competing TTS service on top of our API.
6. Reporting
If you believe an Orpheus-generated audio file is being used in violation of this policy, please contact us at [email protected] with as much context as you can share.
7. Changes
We may update this policy as new abuse patterns emerge. Material changes will be announced on this page with an updated "Last updated" date.