Terms of Service

Effective date: November 4, 2026. The Service is operated by J3X Solutions LLC ("we", "us", "our"). AudioStem.ai is a product and brand of J3X Solutions LLC. These Terms of Service ("Terms") are an agreement between you and J3X Solutions LLC covering your use of the AudioStem.ai website and services (the "Service"). By creating an account or using the Service you accept these Terms. If you do not accept them, do not use the Service.

1. The service

AudioStem.ai uses machine-learning models to separate audio recordings into stems (vocals, drums, bass, instruments) and to produce related outputs such as mixes, chord and MIDI data, and enhanced audio.

On eligible plans the Service also generates new audio with machine-learning models: whole songs from a description, instrument parts that follow a separated track, and sung vocals from lyrics you supply or lyrics the model writes for you ("AI lyrics"). Generated audio is added to your track as a stem or kept as a track of its own.

Separation and generation are estimation processes. Outputs are machine-generated, can contain artifacts, may not match what you asked for, and are provided "as is". We do not guarantee any particular quality, accuracy, or suitability of any output.

2. Your account

You must be at least 13 years old (or the higher minimum age where you live) to use the Service.

Give accurate information when you register and keep it current. An account belongs to one person: do not share credentials, let others use your account, or create extra accounts to dodge limits or bans. You are responsible for activity under your account and for keeping your sign-in method secure.

We identify your account by your email address. It is your unique identifier on the Service. If you sign in with more than one method using the same verified email (for example, a password and "Continue with Google"), those methods resolve to a single account by design: we link them together rather than create a separate account for the same email, so your files, history, and subscription stay in one place regardless of how you sign in. By using more than one sign-in method on the same email address, you consent to this linking.

Account integrity. You may not create or operate multiple or related accounts to bypass billing, free trials, quotas, tier limits, enforcement actions, or a suspension or ban. We may identify related accounts by signals such as email address, payment instrument, or device. If we reasonably believe related accounts are being used to circumvent these Terms or an enforcement action, we may merge, restrict, suspend, or terminate any or all of them and refuse further service.

3. Your content and rights attestation

You keep every right you have in the audio you upload or link, and in the stems and other outputs we generate from it. We claim no ownership of your uploads or your outputs. You grant us only the limited license we need to operate the Service: to receive, store, process, and deliver your content back to you and to anyone you choose to share it with.

When you submit audio you attest that you own it or hold every license and permission needed to process it. Each job records that attestation. Do not submit audio you have no rights to.

The same applies to text you supply for AI generation. Lyrics, descriptions, and directions you give us for an AI song or for AI lyrics must be yours or licensed to you. When you submit your own lyrics (or a reference track) you attest to that, and the request records the attestation. Lyrics the model writes for you, and the songs, parts, and vocals it generates, are outputs: you keep every right you have in them, we claim no ownership of them, and you are responsible for making your use of them lawful.

You decide how you use your stems, and you are responsible for making that use lawful. Separating a recording gives you no rights in the underlying composition or recording that you did not already have.

4. Your voice ("My voice")

Where offered, you can create a voice profile from a short recording of your own singing or speech and use it to have the Service sing AI lyrics, or sing back a vocal stem from one of your own tracks, in that voice ("My voice"). You may create a voice profile of your own voice only. Do not record, upload or otherwise submit another person’s voice, and do not use the Service to imitate a real person, living or dead, or to pass off generated audio as anyone’s genuine performance.

Before a voice profile can be used you must read a phrase we show you, live, into your microphone, and confirm that the voice is your own, that you are at least 18 years old, and that you agree to our processing of your voice as described in the Privacy Policy. Our check compares what was read with the phrase, and the reading with your sample. It records your consent and deters casual misuse. It does not verify your identity, does not prove that the account holder is the person in the recording, and cannot stop a determined attempt to defeat it. Your confirmations are your own statements, and you are responsible for them.

You must be at least 18 years old to create or use a voice profile.

We may disable a voice profile, and remove material made with it, at any time and without prior notice — including on a credible report that the voice is not yours or that it was used to impersonate someone — and misuse of this feature is grounds for enforcement under Section 6.

Audio generated with your voice profile is an output under Section 3: you keep every right you have in it, and you are responsible for making your use of it lawful. Singing in your voice uses your HD and AI minutes. As with other AI generation, a request that delivers usable audio is charged in full even if part of it fails, and a request that delivers nothing is refunded automatically.

Stock voices. Where offered, the Service can also sing a vocal stem again in a stock voice of the other gender. The stock voices are synthetic, generated by machine-learning models from a fixed verse; they are not recordings of, and are not intended to resemble, any real person, and you cannot supply a voice for this feature. A stem sung in a stock voice is an output under Section 3 and is charged and refunded as described above for AI generation. Do not use it to imitate a real person, living or dead, or to pass off generated audio as anyone’s genuine performance.

5. Acceptable use

Do not use the Service to infringe anyone’s rights or to process unlawful content. Do not abuse the platform: no bypassing tier limits, quotas, rate limits, or technical controls, no reverse engineering of the Service or its models, no scraping or automated bulk access, no reselling or white-labeling access without our written agreement, no fraudulent payments, no malicious files, and no interference with other users or our infrastructure. Section 6 describes how we enforce these rules.

No automated access. The Service is for humans using it through the interfaces we provide. You may not access it with robots, crawlers, scrapers, headless browsers, or scripts, and you may not build, offer, or operate an unofficial API, wrapper, or automated integration on top of the Service (including submitting jobs, fetching stems, or driving accounts programmatically) without our prior written permission. If you want programmatic access, ask us at support@audiostem.ai.

Conduct toward us, our staff, and our processes. Do not harass, threaten, defame, or abuse our staff or other users; make bad-faith, coercive, or extortionate demands; manipulate or misuse our billing, refund, chargeback, or dispute processes (including filing chargebacks for charges you authorized, or requesting refunds in bad faith); or attempt to obtain non-public logs, security details, or other internal information outside the channels we provide. We treat abuse of our people and processes as seriously as abuse of the platform, and it is grounds for enforcement under Section 6.

6. Enforcement

We may suspend or terminate accounts that violate these Terms. Violations include: infringing copyright or other rights, circumventing quotas or rate limits, reselling access, scraping, automated or robotic access, operating an unofficial API on top of the Service, attempting to disrupt or probe the Service, and fraudulent payments or bad-faith chargebacks. Repeat copyright infringement is handled under our DMCA Policy, which provides for termination of repeat infringers in appropriate circumstances.

Where practical we warn you first and give you a chance to fix the problem. We reserve immediate suspension or termination, without prior notice, for serious abuse. If we terminate your account for abuse, fees already paid for the current billing period are not refunded.

We may take these actions in our reasonable discretion. For abusive, threatening, or bad-faith conduct we may also restrict how you may contact us (for example, to written channels only or to a designated point of contact), in addition to, or instead of, suspension or termination.

7. YouTube and other linked sources

On eligible plans you can submit a link instead of a file. We fetch linked audio only on your instruction and only for your own processing. You are responsible for having the rights to that audio and for complying with the source platform’s terms of service. If a platform blocks or restricts a fetch, we cannot guarantee the job will complete.

In Create song (Advanced) on eligible plans you can also use a YouTube link as a reference sound. We fetch that audio on your instruction and store it in your account, where it counts toward your storage cap. We use about 30 seconds of it only to shape the sound of songs you create; we never play it back, share it, or let you download the stored copy. It is kept while any of your songs uses it or until you remove it, and deleting the last song that uses it deletes it. We may remove a stored reference, and block the video for every account, in response to a copyright complaint.

Tracks created from a linked source (including YouTube) cannot be published as public share links. The 24-hour share feature is available only for audio you upload yourself. Because linked-source processing depends on the source platform, its availability and behavior may change, be limited, or be withdrawn at any time in line with that platform’s policies, and we may modify or discontinue this capability without notice.

YouTube fetching can stop working at any time. It depends on YouTube, which we do not control and which can block, slow down or change access without warning, so a fetch may fail, be delayed, or be unavailable for hours, days or longer, and we may also pause or turn it off ourselves. It is a convenience, not a guaranteed part of any plan: the number of YouTube fetches a plan lists is the most you can use, not a promise that fetching will work. An outage is not grounds for a refund or a credit, except where the law requires it. You can always upload a file you hold the rights to instead.

8. Subscriptions and billing

Paid plans are billed in advance through Stripe, monthly or annually, and renew automatically until cancelled. Prices and tier limits are shown before you buy; we may change them with notice, effective at your next renewal, except HD and AI minutes, which may change at any time as described below.

Plan changes are upgrade-only. When you upgrade, the price difference for the rest of your billing period is prorated and charged right away, and your renewal date stays the same; the upgrade takes effect once that charge succeeds. If the charge fails, you stay on your current plan until payment goes through. We do not offer downgrades. To move to a cheaper plan, cancel your current plan and subscribe to the new one after the paid period ends.

Cancelling stops future charges. You keep your plan’s features until the end of the period you already paid for. After that your account is paused: for 30 days you can reactivate a plan from the site (we email you about this), or delete your account. If you do nothing, your stored content (uploads, stems, and other outputs) is permanently deleted when that window ends, as with failed payments (Section 9). Your account itself is not deleted: it remains as a free account that you can keep using, or delete at any time from the Account page.

Tier quotas (minutes, storage, daily clips) are fair-use limits, not guarantees of availability.

HD and AI minutes. HD processing (such as audio super resolution and mastering) and the AI features (such as AI songs, AI instruments, AI lyrics, the AI chorus and the voice features) run on costly GPU computing and use your HD and AI minutes. To protect the Service from losses, we may change any of the following at any time, with or without notice: how many minutes each of these features uses (for example per second of audio, or a higher rate for a heavier feature), the minutes included with a plan each month, how many of them the AI features may use, and any other limit on these features. A change applies to requests made after it takes effect; minutes already charged for a request are not recalculated. Minutes you bought in a pack stay in your balance, but what each minute covers may change in the same way.

One-time purchases. Where offered, a track download unlock is a single payment (shown before you buy, currently US$0.99) that opens MP3 downloads of the mix and every stem of ONE track on the Free plan for a fixed period (currently 24 hours). The period starts when your payment succeeds, not when you first download, and it ends early if you delete the track or your account before it runs out. An unlock is not a subscription and nothing renews. Buying an unlock may also keep your other Free tracks from the scheduled clearing for the same period; that keep is a courtesy, not storage you have bought. An unlock may come with a small number of extra uses of Free-plan features (for example an additional YouTube fetch or stem separation). Those credits apply only while you are on the Free plan, have no cash value, cannot be transferred or exchanged, and are not refunded for any reason. If you ask us to open the downloads immediately (which the checkout does), you acknowledge that any statutory right to cancel the purchase ends once they are open.

9. Failed payments

If a renewal payment fails we retry the charge and email you. If payment still cannot be collected, your account is locked and a 30-day notice period begins, with reminders along the way. If the balance is still unpaid when that period ends, your stored content (uploads, stems, and other outputs) is permanently deleted; your account itself remains as a free account unless you delete it. Update your payment method at any time before then to restore access and keep your files.

10. Refunds

Except where the law requires otherwise, payments are non-refundable. At our discretion, we may refund a charge if you ask within 14 days of it and have made no meaningful use of the paid features in that time. Ask at support@audiostem.ai.

A track download unlock (Section 8) is refunded automatically, in full, only when the track was deleted before the downloads opened. Once the downloads have opened, the payment stands even if you delete the track or let the period run out without downloading, and the credits that came with it are never refunded.

11. Storage limits and retention

Some accounts receive a larger storage allowance than their plan lists (for example, an account that has stored a large library on the Studio plan). Such an allowance is attached to the account: it applies while the account is on a plan it was granted for, is kept if the subscription ends and resumes later, and is removed only when the account is deleted or when we withdraw it for misuse.

Each tier includes a storage cap and retention rules. Files may expire automatically (for example, Free-tier artifacts have a limited lifetime) or be removed when you exceed your cap. The Service is a processing tool, not a backup service: keep your own copies of anything you cannot afford to lose. We are not liable for content removed under these rules.

12. Service changes

The Service evolves. Features, machine-learning models, and plan limits may change, be added, or be withdrawn at any time without notice. Where a change removes a material feature of a paid plan, we make reasonable efforts to notify affected subscribers in advance. Model updates can change how outputs sound; we do not guarantee that re-running a job will produce the same result.

13. Our intellectual property

Everything that makes the Service run (the software, the models and their configuration, the site design, and the AudioStem.ai name and branding) belongs to us or our licensors. These Terms give you a personal, revocable right to use the Service and nothing more.

14. Disclaimer of warranties

THE SERVICE IS PROVIDED "AS IS" AND "AS AVAILABLE" WITHOUT WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT. WE DO NOT WARRANT UNINTERRUPTED OR ERROR-FREE OPERATION, OR THAT ANY OUTPUT WILL MEET YOUR REQUIREMENTS.

15. Limitation of liability

TO THE MAXIMUM EXTENT PERMITTED BY LAW, WE WILL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, OR ANY LOSS OF PROFITS, DATA, OR GOODWILL. OUR TOTAL LIABILITY FOR ALL CLAIMS RELATING TO THE SERVICE IS LIMITED TO THE AMOUNTS YOU PAID US IN THE 12 MONTHS BEFORE THE CLAIM AROSE (OR US$50 IF YOU PAID NOTHING).

16. Indemnification

You will defend and indemnify us against claims, damages, and costs arising from content you submit, your use of stems or other outputs, or your breach of these Terms, including any claim that you lacked the rights to audio you processed.

17. Copyright complaints

We handle copyright notices under our DMCA Policy and terminate the accounts of repeat infringers.

18. Termination

You can stop using the Service at any time and delete your account from the Account page; deletion permanently purges your uploads, stems, and profile. We may suspend or terminate your account as described in Section 6, if the law requires it, or if we discontinue the Service (with notice where practical). Sections that by their nature should survive (including content responsibility, disclaimers, liability limits, and indemnification) survive termination.

19. Changes to these terms

We reserve the right to modify these Terms at any time, and we will provide notice of any material changes by posting the updated Terms on our website or within the AudioStem platform. Your continued use of the AudioStem platform following any changes constitutes your acceptance of the updated Terms.

20. Severability and waiver

If any provision of these Terms is held invalid, illegal, or unenforceable by a court of competent jurisdiction, that provision is limited or removed to the minimum extent necessary, and the remaining provisions stay in full force and effect. Our failure or delay in enforcing any provision is not a waiver of it, and no waiver is effective unless we give it in writing. These Terms, together with the Privacy Policy and DMCA Policy, are the entire agreement between you and us regarding the Service.

21. Governing law

These Terms are governed by the laws of the State of California, USA, without regard to conflict-of-laws rules. The state and federal courts located in Los Angeles County, California have exclusive jurisdiction over disputes arising from these Terms or the Service, and you and we each consent to their jurisdiction.

22. Contact

General questions: support@audiostem.ai. Privacy and data rights: privacy@audiostem.ai. Copyright: dmca@audiostem.ai. Security vulnerabilities: security@audiostem.ai.