// legal

Acceptable Use Policy

Last updated 11 September 2026 · Version 1.0

English translation of the Spanish «Política de uso aceptable» v1.0; if the two versions differ, the Spanish one prevails.

Acceptance

Using our platform means accepting this. You are responsible for all use made through your API keys, including by your employees, contractors and the users of your applications.

Prohibited AI practices

You may not use the platform for practices prohibited by Article 5 of Regulation (EU) 2024/1689: subliminal manipulation or exploitation of vulnerabilities causing significant harm, social scoring, individual predictive policing based solely on profiling or personality traits, untargeted scraping of facial images for recognition databases, emotion inference in the workplace or in educational institutions other than for medical or safety reasons, biometric categorisation to infer sensitive data, and real-time remote biometric identification in publicly accessible spaces for law enforcement outside the legal exceptions.

Unlawful or harmful use

Nor for anything unlawful or harmful: child sexual abuse material —zero tolerance, with reporting to the authorities where appropriate—; chemical, biological, radiological or nuclear weapons or explosives; malware, exploits or intrusion into other people's systems —security testing on your own systems is fine—; harassment, threats or incitement to hatred and violence; deliberate disinformation capable of causing harm, non-consensual deepfakes or impersonation of people and organisations; infringement of third-party rights; automated decisions with legal or similarly significant effects on individuals —credit, employment, insurance, benefits, admissions— without the human oversight required by Article 22 of the GDPR; and medical, legal or financial advice presented to a consumer as definitive without professional review.

Abuse of the platform

Nor to abuse the platform: removing or hiding watermarks and provenance information from artificially generated content; circumventing the models' safety filters, including jailbreaking and prompt injection for that purpose; reverse engineering, or using the platform to train or improve a competing inference service; getting around technical limits or billing, for example by splitting usage across several accounts; publishing benchmarks of our service without identifying the model version and methodology, or without our authorisation; and using the models against their own licences, which apply in addition to this policy.

Minors

And we do not offer the service to minors. The platform is for businesses and professionals: do not make it available to consumers who are minors or integrate it into applications directed at minors without the safeguards required by law, and do not send us personal data of children under fourteen without consent and safeguards.

Keys and resale

About keys and resale. Your keys are yours and are used from your infrastructure, server-side: do not hand them to third parties, do not share them between different companies and do not leave them exposed in client-side code or in repositories. You may integrate the platform into your own applications, products or services and offer them to your end users, provided the calls come from your infrastructure and you are responsible for how they use it. What requires a partner agreement signed with us is reselling direct access to the API as such — and, unlike most of the market, we authorise that and pay you for it.

Fair use

Fair use. Flat-rate plans have a fair use policy per tier, described in the documentation. We look at aggregate usage patterns —volumes, frequencies, latencies—, never at the content of your prompts or of the responses, in line with our zero-logging commitment. If your usage pattern corresponds to a higher plan or to a Dedicated deployment, we tell you and we talk.

Breach and reporting

What happens if this is breached. We may suspend access: immediately in serious cases —Article 5, minors, weapons, malware— and, in all other cases, with prior notice and an opportunity for you to fix it. Before a non-urgent suspension we try to talk to you. If you detect misuse of your keys or suspect they have been compromised, write to support@helmcode.com as soon as possible. Anyone can report a breach to that same address, and we treat reports confidentially as far as possible.

Changes to this policy

We may update this policy by publishing the new version with thirty days' notice, or immediately where required by law or for security reasons.

Related documents: Inference Services Agreement (EUSA) · Data Processing Agreement and list of subprocessors · SLA per plan · Partner Agreement.