LEGAL

Terms of Service

These Terms of Service ("Terms") are a legally binding agreement between Sria Digital B.V. ("Xavira", "we", "us", or "our") and you ("Customer", "you" or "your"). They govern your access to and use of the Xavira AI API and portal at xavira.ai and api.xavira.ai (collectively, the "Services").

Effective date: 2026-09-21 · Version: 1.4 · Last updated 2026-09-21
Enterprise agreements. Enterprise customers may negotiate a counter-signed variant of these Terms; contact hello@xavira.ai.

By clicking "I accept", by signing up via our portal, or by accessing the Services in any way, you agree that you have read, understood, and agreed to be bound by these Terms, the Privacy Policy, and (where applicable) the Data Processing Agreement. If you do not agree, you have no permission to use the Services.

These Terms contain disclaimers limiting our liability, an indemnification obligation, and a forum-selection clause (Section 13). Please read carefully.

1.Definitions

2.Eligibility and Account

2.1 Eligibility

You must be at least 18 years of age, or the age of majority in your jurisdiction (whichever is greater), and have the legal capacity to enter into a binding contract. If you are accepting these Terms on behalf of an entity, you represent that you have authority to bind that entity. The Services are intended for business customers building products; individual personal-use accounts may be suspended at our discretion.

2.2 Account creation

To use the Services you must create an Account via the portal using email + password or a supported OAuth provider (Google, Apple). You are responsible for the accuracy of registration information and for keeping it current.

2.3 Credentials

You are responsible for the confidentiality of your portal login credentials and all API keys we issue to you. Every API request authenticated with your key is treated as authorised by you. Rotate keys immediately via the Profile page if you suspect a leak. Notify us at hello@xavira.ai of any unauthorised use of your Account.

2.4 Suspension and termination by Xavira

We may suspend, terminate, or restrict access to all or part of the Services, without liability, if (a) we receive a binding governmental, judicial, or law-enforcement order; (b) we reasonably believe you have breached these Terms (particularly Section 4); (c) we reasonably believe you are involved in fraudulent, misleading, or unlawful activity related to the Services; or (d) it becomes commercially impractical or unlawful to continue providing the Services.

3.Licence Grant

Subject to your compliance with these Terms, Xavira grants you a limited, non-exclusive, non-transferable, non-sublicensable, revocable licence to (a) access and use the Services and the Documentation for your internal business purposes, and (b) integrate the API into your own product or service. All rights not expressly granted are reserved.

You may not (i) resell or rebrand the Services as your own ML infrastructure offering; (ii) reverse engineer, decompile, or attempt to derive the source code of the Services or Models; (iii) extract or scrape Model weights, embeddings, or architectures via probing requests; (iv) interfere with the Services' security, integrity, or rate limits; or (v) use the Services to develop a competing ML inference API.

4.Acceptable Use

You may use the Services to generate adult content for adult-only audiences, subject to the prohibitions below. Adult content (consensual sexual imagery between adults) is allowed by design; the API is built for it.

4.1 Prohibited content

Generation requests for any of the following are blocked at the moderation layer and may result in account suspension or termination. The list is non-exhaustive; see the Content Moderation Policy for procedural detail and updates.

4.2 Prohibited use

You will not, and will not permit any other person to:

4.3 Customer obligations when reselling to end-users

If you re-expose the Services to end-users through your own product, you are independently responsible for:

You may reference Xavira AI factually as a sub-processor or vendor in your privacy policy. You may not represent Xavira AI as a partner, endorser, or co-marketer of your product without our prior written consent.

4.4 Reference images

Supplying your own Reference Images (rather than building a Character from text traits alone) is off by default and opens for your Account only after a credit pack has been purchased; you then enable it from the portal. We may decline or withdraw it for any Account at our discretion. Where it is open for your Account, supply is additionally governed by the Reference Image Supply Agreement, which you must sign before your first supply. Where that Agreement and these Terms conflict on the subject of Reference Images, the Agreement prevails.

In summary, and without limiting the fuller text of that Agreement: every Reference Image you supply must be wholly synthetic (generated by an AI system, not a photograph, recording, edit, or derivative of a real person), must not depict or resemble any identifiable real person — living or deceased, public or private, and regardless of any consent you may hold — and must not depict a person who is, appears to be, or is presented as being under 18. These warranties are given afresh for every image supplied, including images supplied via the API.

We may run automated checks on Reference Images before any Character is created from them, including metadata inspection, perceptual hashing, reverse image search, and classification for apparent age and for resemblance to known public figures. Acceptance of a Reference Image — automatic or by a human reviewer — is not a finding that it complies with this Section, is not a waiver of your warranties, and does not shift any part of your responsibility under this Section or Section 11 to us. We may reject, withdraw, or delete any Reference Image, Character, or Output derived from it at our discretion, including after it has already been used to generate Output.

4.5 Chat

The Chat Service is a tool you build into your own product. Where you offer chat to end-users, you are the provider and operator of that chat experience: you choose the Characters, personas, instructions, audience and context, you decide what your end-users see, and the conversation takes place between your product and your end-user. Xavira is not a party to any conversation, does not monitor conversations in real time, does not review Chat Output before it is delivered, and does not act on your behalf toward your end-users.

Chat Output is machine-generated fiction produced by probabilistic models. It may be untrue, inconsistent, offensive, sexually explicit, emotionally intense, manipulative in tone, or otherwise unsuitable, and a Character may appear to express feelings, intentions, or facts it does not have. Chat Output is never advice (medical, psychological, legal, financial or otherwise) and is not a statement by Xavira.

In addition to Section 4.3, where you offer chat to end-users you are solely responsible for:

Moderation is an aid, not a guarantee. Every chat request is checked by a moderation layer that is included and cannot be switched off: user messages and replies are checked, and verdicts and flag records are returned to you. It is designed around the content rules of payment providers and card networks, but it will not catch everything, it may also block acceptable content, it does not guarantee that you meet any provider's requirements or pass any review, and it does not shift any part of your responsibility under this Section, Section 10 or Section 11 to us. A verdict, a flag, or a suggested reply is information for your own decision, not a finding, certification or advice.

What we keep. We do not store conversations: you send the conversation history with each request. We keep usage metadata (tokens, costs, message counts) and, for any message our moderation flags, that message with a few turns of context, your end-user identifier if you send one, and the verdict. We keep flag records for up to two years (longer where a legal obligation or an open report requires it) for safety, abuse prevention, appeals and legal obligations, and we may report content that indicates child sexual abuse or an imminent threat to life to the competent authorities or organisations (such as NCMEC), and suspend the Account, without notice to you.

5.Customer Data, Inputs and Outputs

5.1 Ownership of Customer Data

As between you and Xavira, you retain all rights, title, and interest, including all Intellectual Property Rights, in and to your Inputs. Subject to the licence in Section 5.2 and to applicable Third-Party Terms, we hereby assign to you all of our right, title, and interest (if any) in and to Outputs generated for you via your authenticated API requests, including the right to use, reproduce, modify, distribute, and commercially exploit them.

Outputs may incorporate elements derived from base models trained on public datasets. We make no warranty that any specific Output is free of third-party rights; you are responsible for any rights clearance required by your downstream use.

5.2 Licence to us

You grant Xavira a worldwide, non-exclusive, royalty-free, sub-licensable (only to Sub-processors as needed) licence to use, store, host, transmit, modify, process, and make available your Customer Data solely to: provide the Services; operate moderation, abuse-prevention, and safety systems; compile Resultant Data; and comply with law.

We do not use your Customer Data to train our Models or any third-party Models without your explicit, separately-obtained consent.

5.3 Resultant Data

Resultant Data is owned by Xavira. We may use it (during and after the term of these Terms) to improve the Services, diagnose issues, publish anonymised benchmarks, and develop new products.

5.4 Feedback

If you provide suggestions or ideas regarding the Services ("Feedback"), you hereby assign to Xavira all right, title, and interest in such Feedback. We may use and commercialise Feedback without obligation to you.

6.Service Provision

6.1 Modifications

We may add, remove, or change Models, features, pricing, rate limits, and any other aspect of the Services. Breaking API changes (removal of an endpoint or required field, removal of a model_id, incompatible response shape change) are announced at least 30 days in advance via email. Additive changes (new endpoints, new optional fields, new models) take effect immediately.

6.2 Beta features

Beta features and models are provided "as is", may be modified or withdrawn at any time, and are excluded from any service-level commitments.

6.3 Pre-alpha service level

Until we publish a formal SLA, the Services are provided on a best-effort basis. We target warm-worker P50 latency of 5 seconds and uptime above 99% during business hours (CET), but make no contractual guarantee. Incidents are listed at status.xavira.ai when published.

6.4 Support

Customer support by email to hello@xavira.ai. Response targets: 1 business day for paying customers, best effort for free-tier and pre-alpha.

7.Fees, Payment, Credits

7.1 Prepaid credit model

The Services operate on a prepaid credit balance ("Prepaid Credits"). You purchase Credits in advance via the Pricing page; we debit them as you use the Services. Per-Model costs are listed at xavira.ai/pricing and in the Documentation.

7.2 No refunds (with limited exceptions)

Prepaid Credits are non-refundable by default. They are non-transferable, do not constitute personal property or currency, and have no equivalent value in fiat currency. Credits become void on termination of these Terms or closure of your Account, except where Section 9.4 provides otherwise.

Chat messages that are blocked by moderation or fail upstream are not charged. Failed generations (HTTP 4xx moderation_blocked, 5xx internal_error / upstream_error, or terminal failed / timed_out states on async gens) are automatically refunded to your Credit balance; no request needed.

For limited exceptions (EU consumer cooling-off, wrongful charges, sustained outage attributable to us, cryptocurrency caveat, chargeback policy) see the full Refund Policy.

7.3 Credit expiration

Each Credit purchase expires twelve (12) months after the date of payment if not consumed. Expired Credits are forfeit and not refunded.

7.4 Payment methods

We do not accept Stripe or major credit-card processors due to adult-content merchant-of-record restrictions. Settlement is via Finby (SEPA bank transfer, BTC, USDT, USDC), PayPal, and any other payment processor we add over time. Payment Processors are independent third parties; their terms and privacy policies apply to your interactions with them.

7.5 Taxes

Fees do not include taxes. You are responsible for all applicable VAT, sales, use, withholding, and similar taxes. Where Xavira is legally required to collect VAT or other tax, we will invoice the appropriate amount and you will pay it.

7.6 Billing disputes

Any dispute regarding Fees must be raised within 30 days of the relevant charge by email with documentation. Our server logs and Postgres ledger are the authoritative record of usage. Continued use after experiencing an issue waives the right to dispute charges related to that issue.

7.7 Suspension for non-payment

If your Prepaid Credit balance is depleted or you fail to pay any invoiced Fees when due, we may suspend access to the Services until the balance is replenished, without liability.

8.Third-Party Offerings

The Services rely on Third-Party Offerings (Supabase, Neon, Netlify, Cloudflare R2, RunPod, OpenRouter, Finby; see Privacy Policy). You acknowledge that (a) the Services' availability depends on those third parties, (b) we are not liable for outages, defects, or data incidents originating with them, and (c) your use of the Services is subject to their applicable terms where they interact with you directly.

9.Term and Termination

9.1 Term

These Terms commence on the earlier of the date you create an Account or first access the Services and continue until terminated.

9.2 Termination by you

You may close your Account at any time by emailing hello@xavira.ai. Closure is effective on receipt of our confirmation reply.

9.3 Termination by us

We may suspend or terminate your Account: (a) immediately for material breach (including Section 4); (b) immediately if required by law or court order; (c) immediately if continued provision becomes commercially impractical; or (d) for convenience with 30 days' email notice.

9.4 Effect of termination

On termination: (a) your right to use the Services ceases; (b) we delete your Customer Data within 30 days, except where retention is required by law (Dutch tax law: 7 years on financial records); (c) unpaid Fees become immediately due; (d) Prepaid Credits are forfeit on termination for material breach; on termination for convenience by us, unused Credits are refunded pro-rata to the original payment method.

9.5 Survival

Sections 1, 4, 5, 7, 9.4, 9.5, 10, 11, 12, and 13 survive termination, along with any other provision that by its nature should survive.

10.Disclaimers

10.1 "As is"

TO THE MAXIMUM EXTENT PERMITTED BY LAW, THE SERVICES, MODELS, AND OUTPUTS ARE PROVIDED "AS IS" AND "AS AVAILABLE" WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT, ACCURACY, OR ABSENCE OF DEFECTS.

10.2 No guarantee of accuracy

Outputs are generated by probabilistic ML models. They may be inaccurate, biased, offensive, or unsuitable for your purpose. We do not guarantee that Outputs accurately reflect real people, places, events, or facts. You must independently evaluate Outputs before relying on them. The Services do not provide medical, legal, financial, or other professional advice.

10.3 Inappropriate Outputs

The Models are capable of generating content that is sexually explicit, suggestive, or otherwise unsuitable for some audiences. While our moderation systems block clearly illegal content (Section 4.1), they are not perfect and may permit Outputs that are inappropriate in your context. You are responsible for downstream filtering and for managing, quarantining, and deleting Outputs as appropriate.

10.4 Third-party content

We do not warrant that Outputs are free of third-party Intellectual Property Rights. Outputs may incorporate stylistic elements derived from training data. You are responsible for rights clearance.

10.5 Chat Output

TO THE MAXIMUM EXTENT PERMITTED BY LAW, XAVIRA ACCEPTS NO RESPONSIBILITY OR LIABILITY WHATSOEVER FOR CHAT OUTPUT OR FOR ANY CONSEQUENCE OF IT, TO YOU, YOUR END-USERS OR ANY THIRD PARTY, INCLUDING ANY EMOTIONAL, PSYCHOLOGICAL, PHYSICAL, FINANCIAL OR REPUTATIONAL HARM, SELF-HARM, RELIANCE ON CHAT OUTPUT, DEFAMATION, OR ANY ACT OR OMISSION OF AN END-USER. YOU USE AND OFFER THE CHAT SERVICE ENTIRELY AT YOUR OWN RISK.

11.Indemnification

You will indemnify, defend, and hold harmless Sria Digital B.V., Xavira, and our affiliates, officers, directors, employees, and agents from and against any claim, loss, damage, liability, cost, or expense (including reasonable legal fees) arising from or relating to:

We will notify you promptly of any claim subject to indemnification. You will take control of the defence with counsel of your choice, subject to our right to participate at our own expense. You may not settle any claim that adversely affects our rights without our prior written consent.

12.Limitation of Liability

12.1 Exclusion of indirect damages

TO THE MAXIMUM EXTENT PERMITTED BY LAW, NEITHER PARTY WILL BE LIABLE TO THE OTHER FOR ANY INDIRECT, INCIDENTAL, CONSEQUENTIAL, SPECIAL, EXEMPLARY, OR PUNITIVE DAMAGES, INCLUDING LOST PROFITS, LOST REVENUE, LOSS OF GOODWILL, OR LOSS OF DATA, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

12.2 Cap on direct damages

TO THE MAXIMUM EXTENT PERMITTED BY LAW, XAVIRA'S TOTAL AGGREGATE LIABILITY ARISING OUT OF OR RELATED TO THESE TERMS, OVER ANY TWELVE-MONTH PERIOD, SHALL NOT EXCEED THE GREATER OF (A) THE AMOUNT YOU PAID US IN THE TWELVE-MONTH PERIOD PRECEDING THE EVENT GIVING RISE TO THE CLAIM, OR (B) ONE HUNDRED EUROS (€100).

12.3 Chat Output

Without limiting Sections 12.1 and 12.2, and to the maximum extent permitted by law, Xavira has no liability of any kind for Chat Output, for conversations between your product and your end-users, or for any act or omission of an end-user (Section 10.5).

12.4 Carve-outs

The limitations in this Section 12 (other than Section 12.3, which protects Xavira only) do not apply to (a) your indemnification obligations under Section 11, (b) your breach of Section 4 (Acceptable Use), (c) either party's breach of confidentiality obligations, or (d) liability that cannot be limited under mandatory law (gross negligence, wilful misconduct, fraud, personal injury, death, or under Article 82 GDPR).

13.Governing Law and Disputes

13.1 Governing law

These Terms are governed by the laws of the Netherlands, excluding its conflict-of-law rules. The UN Convention on Contracts for the International Sale of Goods does not apply.

13.2 Jurisdiction

Subject to Section 13.3, all disputes arising out of or relating to these Terms are subject to the exclusive jurisdiction of the competent courts in Amsterdam, the Netherlands. Each party irrevocably consents to that jurisdiction and waives any objection to forum.

13.3 EU consumer rights

If you qualify as a consumer under EU law, mandatory consumer-protection rules in your jurisdiction may grant you a forum-of-residence right which this Section 13 does not override.

13.4 Informal resolution

Before initiating any formal proceeding, the parties will attempt good-faith resolution by email to hello@xavira.ai for at least 30 days.

14.Copyright Notices

Full notice procedure, designated copyright agent, counter-notification process, and repeat-infringer policy are in the DMCA Takedown Policy. In summary: send a written notice with the elements listed there to hello@xavira.ai with the subject DMCA Notice; we act on valid notices within 48 hours.

For removal requests not based on copyright (depiction without consent, depiction of a minor, defamation, etc.), see the Content Removal Policy.

15.Modifications to These Terms

We may modify these Terms. Material changes are announced by email to the Account address at least 30 days in advance. Minor edits (typos, clarifications, sub-processor list updates) take effect on publication. Your continued use of the Services after the effective date constitutes acceptance.

Each published version of these Terms carries a version number (see the notice at the top of this page). Where a Material change is published, the portal will require logged-in Customers to affirmatively re-accept the new version, with a summary of what changed, at next login — continued use of the API alone does not substitute for that in-portal acceptance for Accounts that log in to the portal.

16.General

16.1 Entire agreement

These Terms, together with the Privacy Policy and (where applicable) the DPA, constitute the entire agreement between you and Xavira regarding the Services and supersede all prior or contemporaneous understandings.

16.2 Severability

If any provision is held invalid or unenforceable, the remainder remains in full force.

16.3 No waiver

Failure to enforce any provision is not a waiver of the right to enforce it later.

16.4 Assignment

You may not assign these Terms without our prior written consent. We may assign these Terms in connection with a merger, acquisition, reorganisation, or sale of substantially all of our assets.

16.5 No agency

Nothing in these Terms creates a partnership, joint venture, employment, or agency relationship between the parties.

16.6 Force majeure

Neither party is liable for delay or non-performance caused by events beyond reasonable control (acts of God, governmental orders, war, terrorism, internet or power outages, Sub-processor outages, labour disputes).

16.7 Notices

Notices to you are sent to the email on your Account. Notices to us must be sent to hello@xavira.ai. Legal process must additionally be served at the registered office of Sria Digital B.V.

16.8 Export and sanctions compliance

You will comply with all applicable export control and economic sanctions laws (EU, US, Dutch, UN). You represent that you are not located in, and will not provide the Services or Outputs to, any party located in or controlled by a country subject to comprehensive sanctions, nor to any party on EU, US, or UN restricted-party lists.

17.Contact

Sria Digital B.V., registered in the Netherlands (Dutch KvK registration on file). General contact and notices: hello@xavira.ai. Privacy and data-subject requests: see the Privacy Policy.