Platform Usage Rules
Effective date: July 24, 2026
Platform and Account Usage Rules
These operational rules govern access to Modelflare accounts, API Keys, credits, payments, models, and infrastructure. They supplement the content and safety requirements below and our Terms of Service. Any exception must be expressly approved by Modelflare in writing and applies only to the approved scope.
Protect Accounts and API Keys
Keep login credentials and API Keys confidential. Do not publish them, embed them in public repositories or client-side applications, or sell, lease, lend, transfer, or share them with unauthorized parties. For team or production use, issue separate keys where practicable, grant only the access required, monitor usage, and rotate or revoke a key promptly if it may have been exposed. You are responsible for activity performed through your account and keys.
Do Not Abuse Registration, Credits, or Payments
Do not create or coordinate duplicate, automated, deceptive, or third-party accounts to evade restrictions or obtain repeated signup credits, referral rewards, promotions, trial access, or other benefits. Do not use stolen or unauthorized payment methods, manipulate payment or usage records, request refunds through deception, or dispute a valid charge solely to avoid paying for completed usage. This does not limit your right to report a billing error or raise a good-faith payment dispute through support.
Respect Access and Service Limits
Do not bypass or attempt to bypass identity, region, model, group, permission, credit, payment, rate, concurrency, request-size, safety, or other service controls. Do not rotate or coordinate accounts, API Keys, IP addresses, devices, integrations, or intermediaries to evade a restriction or regain access that has been limited, suspended, or terminated.
Do Not Attack, Probe, or Disrupt the Platform
Do not scan, probe, test, reverse engineer, overload, interfere with, or attempt unauthorized access to Modelflare, its infrastructure, or its service provider providers. Security testing and vulnerability research require prior written authorization and must remain within the approved scope. Good-faith security reports should be sent through support without accessing, retaining, or disclosing data beyond what is necessary to describe the issue.
Do Not Resell Raw Access or Extract Protected Systems
You may integrate Modelflare into your own product for authorized end users, subject to these rules. Unless a written agreement expressly permits it, you may not resell raw account or API Key access, operate an undisclosed shared-key proxy, or use the service to extract system prompts, hidden rules, model weights, training data, or response patterns, or to train, fine-tune, distill, benchmark, or improve a competing model or model-routing service.
Take Responsibility for Downstream Use
If you connect Modelflare to an application, agent, workflow, plugin, or customer-facing service, use appropriate authentication, key isolation, permissions, rate limits, human review, AI disclosure, content safeguards, logging, and abuse response. You remain responsible for your users, integrations, and automated actions, including activity performed without a person reviewing each request.
Ask Before Exceptional or Unclear Use
If a planned use could conflict with these rules, contact support before proceeding. Silence, technical availability, a successful request, or the absence of an immediate block does not constitute permission. We may investigate suspected abuse and apply the measures described in the enforcement section below, while treating good-faith billing, support, and security reports separately from abuse.
These Platform Usage Rules include Modelflare's Acceptable Use Policy, or AUP, and apply to anyone who accesses Modelflare or submits Input to our products or services, including through direct access, API Keys, customer applications, authorized integrations, or other permitted access methods. We refer to all such people as "Users".
This policy is designed to help Users stay safe and to promote responsible use of Modelflare services. It applies in addition to our Terms of Service, Supported Countries and Regions, pricing, documentation, and any other policies we publish.
As technology, product behavior, and risks evolve, we may update this policy. To enforce this policy, Havenbyte LLC may use detection, review, abuse monitoring, rate limits, access controls, filtering, and other safety mechanisms. If a User violates this policy, we may limit, suspend, or terminate access, and we may block, filter, modify, or refuse Inputs, Outputs, accounts, API keys, service routes, requests, or related activity that violates this policy.
These policies are written in English. If we provide a translation, the translation is for convenience only, and the English version controls to the extent of any conflict.
1. Universal Usage Standards
These universal standards apply to all Users and all use cases.
Do Not Violate the Law or Engage in Illegal Activity
Do not use our services to engage in, facilitate, or encourage illegal activity, including trafficking in controlled substances, human trafficking, prostitution, intellectual property infringement, abuse, exploitation, harming or attempting to harm children, or other activity that violates applicable laws or regulations.
Do Not Harm Critical Infrastructure
Do not use our services to harm, disrupt, degrade, or interfere with critical infrastructure, including telecommunications, healthcare, energy, transportation, water supply, financial services, election infrastructure, medical devices, financial markets, military bases, or similar systems.
Do Not Harm Computer or Network Systems
Do not use our services to develop, deploy, assist, or automate unauthorized vulnerability exploitation, social engineering, malware, ransomware, denial-of-service attacks, botnet tools, unauthorized interception of communications or device monitoring, persistence mechanisms, automated large-scale compromise of computer systems or networks, or methods to bypass security controls.
Do Not Develop or Design Weapons
Do not use our services to develop, design, produce, test, market, or unlawfully obtain weapons, explosives, or dangerous materials. Do not use our services to provide guidance on weaponization, delivery mechanisms, evasion of safety controls, procurement of precursors, or improving weapon capabilities.
Do not use our services to develop military applications, surveillance technologies, or other technologies intended to harm humans, including the design, development, production, or testing of nuclear, biological, chemical, or radiological weapons, military-grade cyber capabilities, missiles, drones, or autonomous weapon systems.
Do Not Incite Violence or Hate
Do not use our services to incite, promote, praise, or glorify extreme violence or identity-based violence against individuals or groups. Do not use our services to promote, praise, facilitate, or advocate hatred against protected groups, or to engage in discrimination, intimidation, dehumanization, or violence based on protected characteristics.
Do Not Violate Privacy, Publicity, or Identity Rights
Do not use our services to engage in, facilitate, incite, or cause stalking, harassment, bullying, or other forms of abuse. Do not use our services to compromise or violate people's privacy, publicity, or other legal rights through illegal or unauthorized means.
This includes creating systems for identifying people in private or restricted spaces, or collecting, sharing, obtaining, or using private information, contact information, health information, biometric information, neural data, confidential information, or other personal data without authorization or another lawful basis.
Do Not Harm Child Safety
Do not use our services to create, obtain, transform, or distribute content that abuses, exploits, or sexualizes minors. This includes child sexual abuse material (CSAM), grooming, sextortion, advice for hiding child abuse, roleplay or descriptions that place minors in sexual contexts, or any content that praises, glorifies, or encourages sexual abuse of children.
Do Not Create Psychologically or Emotionally Harmful Content
Do not use our services to automatically generate or distribute content designed to cause psychological or emotional harm, including content that promotes self-harm, suicide, eating disorders, unhealthy exercise, unattainable body image, bullying, harassment, abuse, humiliation, intimidation, or targeted harassment.
Do Not Create or Spread Harmful Misinformation
Do not use our services to create or spread, at scale, false, inaccurate, misleading, or deceptive content that may cause economic, political, social, or public-health harm. This includes deceptive information about groups, entities, or individuals, and deceptive information about public-health emergencies, civic processes, laws, regulations, procedures, standards, or political or civic leaders.
Do Not Undermine Democratic or Civic Processes
Do not use our services to disrupt, impair, or undermine fair and open democratic or civic processes. Prohibited activity includes distributing false information, targeted voter suppression, creating false grassroots movements, mass-contacting voters or public officials while hiding automation, or interfering with voting, vote counting, or certification processes.
Do not use our services for political campaigning or lobbying, or to generate campaign materials intended to influence political, judicial, social, or economic outcomes at scale. This includes propaganda outputs, personalized or targeted political persuasion, interactive campaign chatbots, misleading political synthetic media, or voter suppression materials.
Do Not Use the Services for Prohibited Criminal Justice, Censorship, Surveillance, or Law-Enforcement Purposes
Do not use our services for predictive policing, prosecution, sentencing, parole, detention decisions, lie detection or similar applications, face recognition, emotion recognition, social scoring, government censorship, battlefield management, non-consensual communication or location tracking, biometric classification used to infer protected characteristics, mass surveillance, or other prohibited law-enforcement purposes.
Do Not Engage in Fraud, Abuse, or Predatory Practices
Do not use our services to create or distribute fraudulent, false, or misleading content designed to deceive, defraud, or exploit people. This includes scams, phishing, forged documents, counterfeit products, fake reviews, fake engagement, spam, and other deceptive practices.
Do not use our services to impersonate a person, organization, entity, or real public figure in order to deceive, mislead, or defraud others. Do not use our services to conduct or facilitate gambling, payday lending, high-cost credit products, exploitative debt collection, or other exploitative economic activity. Do not use our services for other illegal, fraudulent, deceptive, or spam activity.
Do Not Abuse the Platform
Do not use our services in a way that violates our platform rules, including attempts to undermine safeguards, evade built-in model limitations, extract model weights, system prompts, hidden rules, training data, or response patterns without authorization, conduct jailbreaks or prompt injection attacks, scrape models without authorization, copy models, train, fine-tune, or distill competing models, or violate service limits such as rate limits, concurrency limits, request-size limits, payment limits, or access controls.
Do Not Generate Sexually Explicit Content
Do not use our services to generate sexually explicit content, including content created for explicit sexual purposes or sexual gratification. This includes pornography, explicit sexual acts, fetish content, sexual fantasy, erotic chat, or incest content, whether textual, visual, audio, or otherwise.
This restriction does not prohibit legitimate sexual and reproductive health, scientific, educational, or artistic use cases, provided they comply with applicable law and do not involve minors, exploitation, coercion, non-consensual sexual content, or other prohibited content.
2. High-Risk Use Case Requirements
Some use cases present a higher risk of harm because they affect areas that are important to public welfare, individual rights, and social fairness. For these use cases, relevant human expertise should be integrated, and end users should know when AI is involved in generating an Output.
For the high-risk use cases listed below, Customer must implement these additional safeguards.
Human in the Loop
When using our products or services to provide advice, recommendations, assessments, or subjective decisions that directly affect an individual or consumer, a qualified professional in the relevant field must review the content or decision before it is distributed, finalized, or relied upon. Customer or Customer's organization is responsible for the accuracy, suitability, and appropriateness of the information.
Disclosure
If model Output is presented directly to an individual or consumer, Customer must disclose that AI is being used to help generate the advice, decision, recommendation, or content. This disclosure must be provided at least at the beginning of each relevant session or interaction.
High-Risk Use Cases
High-risk use cases include:
- Legal: use cases involving legal interpretation, legal guidance, or decisions with legal effects.
- Healthcare: use cases involving healthcare decisions, medical diagnosis, patient care, treatment, mental health, or other medical guidance. General wellness advice, such as sleep, stress, nutrition, or exercise tips, is not included unless it becomes medical guidance.
- Insurance: use cases involving health, life, property, disability, or other insurance underwriting, claims processing, coverage decisions, or eligibility decisions.
- Financial: use cases involving financial decisions, including investment advice, loan approval, financial eligibility, or creditworthiness.
- Employment and Housing: use cases involving decisions about a person's employability, resume screening, hiring tools, other employment decisions, rental eligibility, housing eligibility, or mortgage eligibility.
- Academic Testing, Certification, and Admissions: use cases involving standardized testing companies that administer school admissions, language proficiency, or professional certification exams, including assessing, scoring, or ranking prospective students, or organizations that evaluate and certify educational institutions.
- Media or Professional News Content: use cases involving automated generation of content with our products or services and publication of that content for external consumption.
3. Additional Use Case Guidelines
The following guidelines apply whether or not the use case is otherwise classified as high risk.
Consumer-Facing Chatbots and Agents
All consumer-facing chatbots, externally facing AI agents, or interactive AI systems must disclose to Users that they are interacting with AI and not a human. This disclosure must be provided at least at the beginning of each chat session or comparable interaction.
Products Serving Minors
Products that serve minors, including products that allow minors to directly interact with systems powered by our APIs, must implement appropriate age-aware safeguards, parental or guardian consent where required, content safety protections, reporting mechanisms, and compliance controls. Such products must not encourage minors to hide activity from parents, guardians, schools, or trusted adults, and must not expose minors to prohibited sexual, exploitative, self-harm, or otherwise harmful content.
Agentic Use Cases
Agentic systems remain subject to this Usage Policy. Customers are responsible for ensuring that agents, tools, plugins, connectors, automation, memory, retrieval, browsing, code execution, file access, or other delegated capabilities do not perform actions that would violate this policy if performed directly by a human User.
Customers must use appropriate permissioning, logging, review, sandboxing, confirmation steps, or other safeguards for agents that can take actions, access sensitive data, communicate externally, spend money, modify systems, execute code, or affect third parties.
Connectors, Plugins, and Integrations
If Modelflare offers a connector, plugin, tool, or integration directory, listed integrations must comply with the directory rules and this Usage Policy. Integrations must accurately describe their capabilities, requested permissions, data handling, and user-facing behavior.
4. Enforcement
We may investigate suspected violations using automated systems, human review, customer reports, provider reports, payment provider signals, security telemetry, request metadata, diagnostic logs, or other information available to us.
Depending on the nature, severity, and recurrence of a violation, we may take actions including warning the Customer, filtering content, blocking requests, disabling specific models or service routes, limiting rate or concurrency, disabling API keys, suspending accounts, terminating access, refusing refunds related to misuse, preserving evidence, or reporting activity to appropriate authorities where required or permitted by law.
Customers are responsible for their Users and downstream integrations. A violation by a User, customer, employee, contractor, agent, integration, or automated system may be treated as a violation by the Customer.