KurdishTTS — Terms of Use

Version: 2026-10-03 (draft) · Effective: [date of publication] Authoritative language: English. Translations are for convenience; if they differ, this English text prevails.

1. About these terms

KurdishTTS (https://kurdishtts.dks.news, "the service") is a free, non-profit project of the Dijital Kurdistan (DKS) / PezkuwiChain initiative. Its goal is a Kurdish voice for the Kurdish nation. It is operated by Pez Kiwi Comp, Georgia, Sagarejo Region, village Kvemo Lambalo, 4th street N2 ("we", "us").

By using the service you agree to these terms. Our Privacy Policy explains how we handle personal data; it is not part of this contract. If you do not agree, please do not use the service.

2. What the service offers

• Read-aloud. Type Kurmancî or Soranî text (up to 600 characters) and receive audio.
• Voice donation (/bexsh). Adults can sign in with a Pezkuwi wallet, read sentences aloud, and check other donors' recordings. This helps us build our own Kurdish voice.
• API. Programmatic access to speech generation is available only with an API key we issue, at our discretion (for example to dks.news).

The service is free. We may change, limit, suspend or stop any part of it at any time, including for maintenance or security.

3. Non-commercial use and the current voice

The voices currently used for read-aloud are Meta's MMS-TTS model for Kurmancî (facebook/mms-tts-kmr-script_latin) and razhan/mms-tts-ckb for Soranî. The Kurmancî model is licensed under Creative Commons Attribution-NonCommercial 4.0 (CC BY-NC 4.0); the Soranî model is a fine-tune of Meta's MMS and is treated the same way. Both are used non-commercially only. They are a temporary solution until our own voice is ready.

Therefore:

• You may use audio from the service only for personal, educational, journalistic or other non-commercial purposes. You may not sell it, use it in advertising, or use it in a product or service offered for money.
• We grant no commercial rights in generated audio. We do not promise that any audio is free of third-party rights. You are responsible for the text you enter and for how you use the result.
• When you share generated audio, please credit it, for example:

  Voice: KurdishTTS (kurdishtts.dks.news), using MMS-TTS by Meta AI / razhan/mms-tts-ckb, licensed CC BY-NC 4.0 (creativecommons.org/licenses/by-nc/4.0).

When the voice changes (for example, to our own model), we will update this section. The new terms will apply to audio generated after the change.

4. Acceptable use

You must not use the service, and in particular the read-aloud, to:

1. create or spread content that is unlawful, or that incites violence, hatred or discrimination against any person or group (including on grounds of ethnicity, nation, religion, gender or sexual orientation), or that harasses, threatens or defames anyone;
2. impersonate a real person or organisation, or make audio that falsely appears to be said by, or endorsed by, a real person. This includes fraud, scams, fake emergency calls and misleading "voice notes";
3. produce sexual content involving minors, or any content that exploits or endangers children;
4. mislead people in a way that could cause harm. For example, do not present generated audio as a real recording of events, of authorities or of this project;
5. overload, scrape or attack the service; get around rate limits or access controls; or use automated means beyond normal use of the public page. To report a security problem, write to security@pex.mom; good-faith reports are welcome;
6. break the law that applies to you or anyone else's rights.

We may refuse a request, remove cached audio, block access or a wallet address, or report clearly illegal content to authorities.

5. Voice donation

5.1 Who can donate

You must be 18 or older. You must accept the separate consent text shown before your first recording.

5.2 Your promises

By donating you confirm that:

• the voice is your own, and you read only for yourself. You do not record anyone else, read on someone else's behalf, upload someone else's recording, or use a synthetic or altered voice;
• no one else can be clearly heard in the background;
• you read only the sentence shown, and you do not add names, personal information or any other content;
• you sign in only with a wallet you control.

5.3 Permission you give us

Your recordings stay yours. You give us a non-exclusive, worldwide, royalty-free permission to use your recordings, and your reading performed in them, only for:

(a) storing, checking and quality-reviewing them (including by other donors, section 5.4); (b) training and testing DKS/PezkuwiChain's own speech models; and (c) offering and publishing the models that result.

This permission covers any copyright, performer's or related rights you may have in the recordings.

We will not sell your recordings, share them outside the project, publish them, or use them for any other purpose. The details are in the Privacy Policy.

This permission ends for the future when you delete your recordings. It continues only for models that were already trained before you deleted them, which cannot practically "unlearn" a recording. This is explained in the consent text and the Privacy Policy.

5.4 Checking other donors' recordings

When you review recordings:

• listen only to check them. Do not record, download, copy, share or publish other donors' recordings, and do not try to find out who made them;
• vote honestly. Do not coordinate votes or use several wallets.

Breaking this rule ends your access, and may break the law.

5.5 No payment

Donation is voluntary and unpaid. You receive no money, tokens (including HEZ/PEZ) or other reward for donating or reviewing, and you gain no ownership or revenue share in any model.

5.6 Quality checks and removal

Recordings are checked automatically (length, silence, clipping, reading speed) and by other donors. We may reject or delete any recording, and block a wallet address that breaks these terms.

5.7 Leaving

You may stop at any time and delete your recordings with the "Delete my recordings" button. The effect is described in the Privacy Policy, section 7.

6. Your wallet

You sign in by signing a message with your Pezkuwi wallet. We never ask for your recovery phrase or private keys — anyone who does is not us. You are responsible for the security of your wallet. Signing in creates no blockchain transaction and costs nothing.

7. Our content and code

The sentences you read come from Common Voice's sentence collection and are in the public domain (CC0). Our site design, texts and software belong to us or our licensors. The source code is published at github.com/pezkuwichain/pezkuwichain-kurdishtts under the licence stated there.

8. No warranty

The service is offered "as is" and "as available". Speech synthesis makes mistakes: words can be mispronounced, skipped or misread, especially names, numbers and mixed-language text. Do not rely on the service where errors could cause harm — for example for medical, legal, safety or emergency information. We do not promise that the service is uninterrupted, error-free, or suitable for any particular purpose.

9. Limitation of liability

The service is a free, non-profit offering. To the extent the law allows, we are not liable for indirect or consequential loss, loss of data, profits or reputation, or for loss caused by your use of generated audio. Our total liability to you is limited to EUR 100.

Nothing in these terms limits liability for death or personal injury caused by negligence, for intent or gross negligence, for fraud, or for anything else that cannot be limited by law. Nothing limits your rights under data protection law or the mandatory consumer rights of the country where you live.

10. Responsibility for misuse

If you use the service in breach of these terms or the law, you are responsible for the consequences. This includes reasonable costs we incur because of claims by others, to the extent the law allows.

11. Ending access

You can stop using the service at any time. We may suspend or end your access, or block a wallet address, if you breach these terms or if we have to in order to protect the service, other users or the law. Sections 3, 5.3 (for models already trained), 8, 9, 10 and 13 survive.

12. Changes to these terms

We may update these terms. We will publish the new version here with a new date. Changes that reduce your rights take effect no earlier than 30 days after publication. Changes to how donated recordings are used never apply to your recordings without your new consent. If you keep using the service after a change takes effect, the new terms apply.

13. Governing law and disputes

These terms are governed by the law of Georgia. Disputes go to the courts of Tbilisi, Georgia.

If you are a consumer, this choice of law does not take away the protection of mandatory laws of the country where you live, and you may also bring proceedings in the courts there. Data protection complaints can always go to your local supervisory authority (Privacy Policy, section 11).

If any part of these terms is invalid, the rest remains in force.

14. Contact

Pez Kiwi Comp, Georgia, Sagarejo Region, village Kvemo Lambalo, 4th street N2. General and legal notices: privacy@dks.news (subject "Legal notice") · Privacy: privacy@dks.news · Security: security@pex.mom.