Version: May 12, 2026
This English version is provided for reference. The Japanese version is the controlling document; in the event of any conflict, the Japanese version prevails.
These Terms of Service (the “Terms”) govern the use of the translation diagnostic service “CATER v2” (the “Service”) provided by ErudAite, Inc. (“we”, “our”, or “us”). By using the Service, you (the “User”) agree to be bound by these Terms.
The Service is part of the ErudAite product family. Sign-in uses the same ErudAite account as ErudAite Stand By Me (“SBM”), and the SBM Terms also apply to signed-in Users. Where the SBM Terms and these Terms conflict in respect of the Service, these Terms control.
We may update these Terms at any time without prior notice, by posting the updated version on the Service or our website.
2.1 The Service produces scores and diagnostic reports for a pair of (source text, translation candidate) along six axes (GP / SI / FC / TC / DC / CSA). The Service is a diagnostic tool, not a machine translation service, and operates without dependence on a human reference translation. It is grounded in the academic framework “CATER v2: A Strategy-Theoretic Diagnostic Framework” (Iida, 2026).
2.2 The Service is provided as is. Diagnostic output is intended as decision support and not as a substitute for human judgement. Interpretation of and reliance on the Service’s output is the User’s sole responsibility.
2.3 The Service may experience delays, retries, or partial outages caused by load, LLM provider rate limits, or upstream incidents. Stage-level timeouts may require automatic retries.
2.4 We may upgrade the Service without prior notice, which may change scoring methodology, prompts, model selection, the user interface, and operational behaviour. To preserve reproducibility, the orchestrator version is recorded in every report.
3.1 The Service is offered in the following tiers:
3.2 Account creation and sign-in are handled exclusively through the SBM sign-up flow at https://sbm.erudaite.ai/signup?from=cater. The Service has no independent sign-up flow. Age verification, identity verification, password reset, and other account-management procedures follow SBM’s policies.
3.3 We may refuse registration or suspend a User where: (a) the User provided false information; (b) the User has previously violated these Terms or the SBM Terms; or (c) we otherwise consider the use unsuitable.
4.1 Each plan has a monthly quota for Basic and Advanced evaluations and a per-evaluation input character limit, as displayed in the Service. The quota resets on the first day of each calendar month (Japan Standard Time).
4.2 We may change the monthly quotas and per-plan feature allocations without prior notice. Where we reduce a quota, we will use commercially reasonable efforts to provide advance notice.
4.3 Where the User’s SBM plan changes (Free / Starter / Standard / Beyond), the corresponding plan in the Service updates accordingly. Termination or freezing of the SBM account also terminates the Service session.
5.1 Users on Free and above may purchase additional Advanced or Ultimate evaluations one at a time when their monthly quota is exhausted. Guest (non-signed-in) Users cannot make such purchases.
5.2 The price (including consumption tax where applicable) is shown on the Stripe Checkout page and becomes binding upon User confirmation.
5.3 Payment is processed by Stripe, Inc. (United States) via the ErudAite Account Service. The Service itself does not store card numbers.
5.4 The purchased evaluation credit is granted to the User’s ErudAite account immediately upon Stripe approval.
5.5 Because the purchased credit is consumed in providing immediate digital service, refunds (including pro-rata refunds) cannot be granted after Stripe approval. See the Notice based on the Specified Commercial Transactions Act for details.
6.1 All intellectual property rights in the Service, including its theoretical foundation (the CATER v2 framework), diagnostic algorithms, prompts, scoring formulas, UI, logos, and documentation, belong to us or to the parties licensing them to us.
6.2 Intellectual property rights in the source text, translation candidate, Translation Brief, and Reference materials (the “User Content”) remain with the User or the rightful owner.
6.3 Diagnostic reports generated by the Service may be freely stored, quoted, distributed, and modified by the User, provided that no third-party rights are infringed.
6.4 The User represents and warrants that they have all rights, licences, and permissions necessary for us and our processors (including LLM providers, hosting providers, and the payment processor) to handle the User Content as required to provide the Service.
6.5 If you cite the Service’s diagnostic output in academic work, please cite the CATER v2 reference and the orchestrator version printed in the report. This request supports reproducibility but is not a contractual obligation.
7.1 To preserve reproducibility and reduce inference cost on identical inputs, the Service caches per-stage outputs on a hash-keyed deterministic cache. The cache key contains no authentication-derived identifiers (no user ID, no email address, no plan).
7.2 As a consequence, two Users submitting identical input share the same cache entry. The same input always yields the same output, regardless of who submits it or when.
7.3 We do not retain input text beyond what is required to produce the diagnostic report. Cache entries automatically expire after up to 30 days.
7.4 Diagnostic reports themselves are not stored on our servers. Reports are saved only in the User’s browser (localStorage) and may be cleared from the browser settings. Users wishing to retain reports should use the JSON export or PDF print feature.
7.5 We do not use Google Analytics or similar marketing analytics tools.
7.6 For details, see the Privacy Policy.
Users must not engage in any of the following:
8.1 Acts that violate laws or public order and morals;
8.2 Acts related to criminal conduct;
8.3 Disrupting or destroying our servers or network;
8.4 Acts that interfere with the operation of the Service;
8.5 Improperly collecting or storing personal information of other Users;
8.6 Impersonating another User or improperly obtaining or reselling authentication tokens;
8.7 Providing benefits, directly or indirectly, to anti-social forces in connection with the Service;
8.8 Submitting content that violates the ethics policies of the Service or its upstream LLM providers (including child sexual abuse material, content inciting violence or terrorism, requests for medical advice that constitutes the practice of medicine, etc.) for diagnostic evaluation;
8.9 Using the Service’s API as a general-purpose LLM gateway or in a manner that violates the LLM providers’ terms;
8.10 Calling the Service through automated scripts or bots at a frequency materially higher than normal human use;
8.11 Other conduct that we deem inappropriate.
9.1 We may suspend or interrupt all or part of the Service without prior notice in any of the following cases:
9.2 We are not liable for any loss or damage suffered by the User or any third party as a result of such suspension.
10.1 We may, without prior notice, restrict the User’s use of all or part of the Service or terminate the User’s use where: (a) the User violates any provision of these Terms; (b) the SBM account information is found to be false; or (c) we otherwise determine that use is inappropriate.
10.2 We are not liable for any damage caused thereby.
11.1 Our liability for non-performance is excluded except in cases of our intentional misconduct or gross negligence.
11.2 The Service is provided as is, without any warranty of completeness, accuracy, reliability, or fitness for a particular purpose.
11.3 Diagnostic output (scores, error citations, edit suggestions, axis-level evaluations, etc.) carries the inherent uncertainty of large language model outputs. We make no warranty as to its accuracy, reproducibility, fitness for a particular purpose, conformity with industry standards or individual contracts, or freedom from defects.
11.4 The User bears all risk of acting on diagnostic output, including but not limited to: economic loss from accept/reject decisions on translations, disputes arising from disclosing output to third parties, academic responsibility for citing output, and legal/medical/financial harm caused by overlooked errors.
11.5 We are not liable for differences between past and future reports caused by Service upgrades that change prompts, scoring formulas, or model selection. Users requiring reproducibility should note the orchestrator version printed in each report.
11.6 We are not liable for outages, transmission errors, rate limits, or billing errors caused by upstream third-party services (LLM providers, hosting, authentication, payment processors, etc.).
11.7 To the maximum extent permitted by law, we are not liable for any direct, indirect, incidental, or consequential damages arising out of use of the Service, including business interruption, data loss, economic loss, lost opportunity, or loss of goodwill.
We may modify or discontinue the Service without notice and are not liable for any damage caused thereby. To support academic use, we will use commercially reasonable efforts to publish version history (changelog or documentation) for prompt or scoring changes.
We may amend these Terms at any time. Amendments take effect when posted on our website. By continuing to use the Service after the amendment, the User is deemed to accept the amended Terms.
Notices and communications between the User and us are made through methods we designate.
15.1 The User may not assign or pledge their position under the User agreement or any rights or obligations under these Terms to any third party without our prior written consent.
15.2 If we transfer our business in respect of the Service to a third party (including via corporate split or other substantive transfer of business), we may transfer the User agreement, the rights and obligations under these Terms, and the User’s registration data to the transferee. The User consents to such transfer in advance.
16.1 These Terms are governed by Japanese law.
16.2 Any dispute arising in connection with the Service is subject to the exclusive jurisdiction of the court having jurisdiction over our head office.
See our Privacy Policy, which forms part of these Terms. Pay-per-evaluation transactions are also covered by the Notice based on the Specified Commercial Transactions Act (Japan).