Acceptable Use Policy

Effective 11 November 2026

This policy applies to use of the Taldermont website, dashboard and API, including use through an application you build. It forms part of the Terms of Service of TALDERMONT LIMITED. You are responsible for use under your account and for the rules you apply to users of your application.

1 Model and provider rules

Independent providers operate the third-party open-weight models you access through the Service. DeepInfra, Inc. currently provides inference. Follow its applicable terms and the licence and usage rules of your selected model, as shown in the model catalogue. The following links help you find the relevant rules.

Where a provider or model rule is stricter, the stricter rule applies to requests for that model.

2 Safety and individual rights

You may not use the Service to produce, request or distribute content, or to carry out activities, that:

  • Sexually exploit or sexualise children, including by creating or sharing child sexual abuse material. If we become aware of such material, we take appropriate action, including preserving available evidence and reporting it where required by law.
  • Encourage or enable violence, terrorism, violent extremism, trafficking or exploitation, or assist with explosives or biological, chemical, nuclear, radiological or other weapons capable of mass harm.
  • Threaten, harass, bully, stalk or intimidate people, or encourage self-harm or suicide or provide instructions for it.
  • Break laws applicable to your activity or the location where content is used, infringe another person’s intellectual property or privacy rights, or collect, profile, use or disclose personal data without a lawful basis or required safeguards.
  • Enable scams, fraud, phishing, money laundering or other financial crime. This includes accessing the Service in breach of sanctions, export controls or the regional restrictions in the Terms of Service.

3 Impersonation and consequential decisions

Do not impersonate people or organisations or make non-consensual deepfakes of real people. Do not use the Service to run disinformation or covert influence campaigns, create fake reviews or accounts, or organise deceptive coordinated activity. Do not present AI-generated material as human-written in a way that misleads and harms others.

If an AI system interacts with people, tell them they are dealing with AI when that would not be apparent or disclosure is required by law. Apply any legally required labelling for generated or manipulated content. Do not use manipulative or exploitative techniques likely to cause harm. Follow applicable prohibitions and transparency duties, including Articles 5 and 50 of the EU AI Act where they apply.

Do not make automated decisions that significantly affect people without appropriate human oversight and the safeguards required by law. This applies, for example, to employment, credit, housing, insurance, education, access to essential services and legal status. Medical, legal or financial advice provided to others requires qualified professional review where the law or the applicable model or provider policy requires it.

4 System security accounts and Points

  • Do not use the Service to attack a system or network, develop malware or ransomware, or build tools for unauthorised access. Obtain our written permission before testing the security of the Service. Report suspected vulnerabilities to the contact below.
  • Do not overload the Service, send bulk unsolicited messages, or bypass limits on spending, billing, requests or access. Do not evade a model safeguard to obtain content prohibited by this policy.
  • Keep API keys confidential. Do not sell, rent or publicly share keys or accounts, resell gateway access, or offer raw access to the Service. You may build an application that uses the Service in accordance with the Terms.
  • Do not transfer or sell Points, try to convert them to cash outside the authorised refund process, or open multiple accounts to obtain promotional Points.
  • Follow restrictions on using outputs to develop competing models when an applicable provider term or model licence prohibits that use.

5 Applications you build

If other people use your application, apply rules at least as protective as this policy. Give them a way to report abuse and act on reports. Provide the privacy notices and AI disclosures required for your application, and secure any permissions and legal basis needed to transmit data to Taldermont and the relevant Model Provider. Put any required data processing terms in place before submitting personal data.

6 Reports enforcement and appeals

We do not routinely read or pre-screen request content. We may act on reports, unusual usage metadata, or information from a Model Provider, payment provider or authority. Depending on the circumstances, we may warn you and ask you to fix the issue, block a key or model, restrict or close an account under the Terms, or report unlawful content or activity to the authorities. Where practicable and lawful, we explain what we did and why. We may limit what we share to protect security, an investigation or a legal obligation.

To report abuse or unlawful content, give us the location or relevant request reference if known, the reason for your concern and any useful supporting information. To appeal a restriction, send your account email, identify the action and explain why you believe it should be reconsidered. We prioritise reports and appeals by urgency and respond within a reasonable time. The Terms of Service and Refund and Withdrawal Policy govern any remaining paid Points if your account closes.

Contact

TALDERMONT LIMITED
Hong Kong Business Registration Number 81573444
Unit 704C, 7/F, Block 3, Nan Fung Industrial City, 18 Tin Hau Road, Tuen Mun, N.T., Hong Kong
For abuse reports, security and appeals, email info@taldermont.com