Terms & Conditions

§ 1 Who We Are and What These Terms Cover

These Terms of Service ("Terms") govern your access to and use of the websites, applications, models, application programming interfaces (APIs), chat interfaces, and online services made available by Heysuvi Labs, LLC, a Wyoming limited liability company with its registered business address at 7345 W Sand Lake Rd, Ste 210, Orlando, FL 32819, USA ("Heysuvi," "we," "us," or "our"), as described in § 2 (collectively, the "Service").

By creating an account, obtaining API credentials, accessing, or using the Service, you ("you," "User") agree to these Terms and to our Privacy Policy (see § 15). If you do not agree, do not use the Service.

These Terms are a complete, standalone agreement and replace in full the prior "Appendix to the TTI – Technologie-Transfer-Initiative GmbH an der Universität Stuttgart General Terms and Conditions" and any other terms previously governing the Service.

§ 2 The Service

Heysuvi is an AI research lab. The Service may include, in whole or in part and from time to time:

• our websites (including greenai.eco and heysuvi.com), research outputs, publications, and demonstrations;
• GreenAI (including version 1.0.1 and successor versions), our energy-efficient multimodal / language model product for private and CPU-oriented workflows;
• the GreenAI API, through which authenticated callers submit prompts and related inputs for inference in a chosen hosting region (see § 4 and § 5);
• a future user chat interface (web or app) that lets end users converse with GreenAI, subject to the same Terms and Privacy Policy when offered;
• other AI models, tools, or interfaces through which you may submit inputs (including documents and other data) for processing; and
• learning and certification offerings (courses, assessments, and digital certificates), where and to the extent we choose to offer them.

We may offer some features only intermittently, and we may add, change, suspend, or discontinue any feature at any time. Where a change is material and adverse to you, we will give reasonable notice under § 20. Certificates already issued remain valid (see § 13).

The Service relies on various artificial-intelligence models. The specific model versions, endpoints, and infrastructure used may change from time to time. Availability of the EU region, chat UI, or particular API features may roll out after the US API MVP.

Who may use GreenAI. The Service is available to anyone with a valid subscription (the general public / all paying users). Assistive student chat and tutoring-style Q&A is our primary intended use case, not a restriction on who may use GreenAI.

Primary intended use. GreenAI’s primary product use case is assistive student chat and tutoring-style question-and-answer (including via the API and, when offered, a user chat interface)—useful for students and other learners, but not a students-only or exclusive-audience product. GreenAI is designed to help users explore concepts, practice understanding, and find information. It is not offered as a system that makes or replaces institutional academic or admissions decisions.

Features not offered as GreenAI product functions. Without limiting our right to change the Service, GreenAI as described in these Terms does not include, as product features: (a) automated grading or scoring that determines academic credit or formal assessment outcomes; (b) admissions, eligibility, or similar selection decisions about natural persons; or (c) biometric identification, emotion recognition, or camera-/microphone-based exam proctoring. Separate Heysuvi learning or certification features, if offered, are described in § 13 and are distinct from GreenAI student chat.

§ 2a AI System Transparency (EU AI Act)

GreenAI is an artificial intelligence system. When you (or an end user of an application you build) interact with GreenAI chat or receive generative outputs, you are interacting with AI, not a human.

Where the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) requires transparency for AI systems that interact directly with natural persons (including Article 50), we design GreenAI chat and the GreenAI API so that users are informed of the AI nature of the interaction in a clear and distinguishable way—at or before the first interaction in our chat interface when offered, and via API metadata and documentation that instruct API customers to surface an equivalent notice to their end users.

API customers who present GreenAI to natural persons must inform those persons that they are interacting with an AI system and that outputs may be incorrect, unless a lawful exception applies. Burying disclosure only in these Terms is not sufficient for end-user-facing chat.

Machine-readable indications that content was AI-generated may be provided where technically feasible (for example, response headers or metadata).

§ 3 GreenAI Product, Method, and Research Citations

Product. GreenAI is a product of Heysuvi Labs, LLC. Public product identity refers to GreenAI by Heysuvi Labs (for example, GreenAI 1.0.1). Marketing and product UI do not claim affiliation with third-party model brands (see § 11).

Method. GreenAI may be described, factually, as using a ternary neural network (ternary-net) method for efficient inference. That description is technical, not a trademark claim about any third party.

Research citations. Related research associated with Heysuvi / Nakul Vyas that may be referenced in connection with the Service includes:

• Nakul Vyas, SUVI: Scalable Unified Vector Intelligence for Efficient Edge Deployment, IEEE Conference on Artificial Intelligence (CAI), 2026, https://doi.org/10.1109/cai68641.2026.11536337.
• Nakul Vyas (and co-authors if any), Intervention-Screened Visual Perception for Small On-Device Agents, short paper / presentation associated with the NeurIPS 2026 Workshop on Small Language Models for Agentic Systems (SLM-Agents / SLM Workshop 2026, NeurIPS Satellite, Paris); public PDF linked from the Heysuvi site dataroom: https://openreview.net/pdf?id=kMETBBIrP0.

Citations are informational. They do not expand or shrink your license to use the Service beyond these Terms.

§ 4 Dual-Region Hosting (United States / European Union)

GreenAI API (and, when offered, chat) is designed to run in two hosting regions. Callers choose a region by using the corresponding regional base URL (not by a per-request header or query switch on a shared global front):

• United States: Azure region eastus (East US), typically via a hostname of the form us.api.<domain>.
• European Union: Azure region swedencentral (Sweden Central), typically via a hostname of the form eu.api.<domain>, when that region is provisioned and publicly available.

Data residency implications. Inference requests (prompts, context, and related operational metadata needed to serve the request) are processed in the region corresponding to the base URL you call. Choosing the US endpoint means processing in the United States; choosing the EU endpoint means processing in the EU (Sweden Central) when that stack is live. There is no silent cross-region failover: if your chosen region is unavailable, requests are not automatically redirected to the other region.

Account, billing, support, website analytics, and certain back-office systems may still be operated from or involve providers in other countries, as described in the Privacy Policy. Regional choice for inference does not by itself create a separate privacy regime; it determines where GreenAI inference for that call is executed.

Until the EU stack is live, only the US endpoint may be available. Attempting to use an unprovisioned EU hostname may fail to resolve or return an unavailable response.

§ 5 API Access, Authentication, and Rate Limits

Credentials. Access to the GreenAI API requires authentication credentials we issue (for example, API keys or tokens). You must keep credentials confidential, use them only for your own authorized use, and not share or embed them in public client-side code. You are responsible for all activity under your credentials.

Rate limits and fair use. We may apply rate limits, quotas, concurrency caps, and abuse controls. Limits may differ by plan, region, or endpoint and may change with notice or as documented in API documentation. You may not attempt to circumvent rate limits (including by creating multiple accounts or keys without authorization).

Guardrails. To protect users and the Service we may apply authentication requirements, rate limits, concurrency caps, payload size limits, tool allowlists (including limiting server-side tools such as web search), abuse and jailbreak controls, logging redaction, and regional endpoint pinning, as described in our API documentation. Circumventing those controls is prohibited under § 9.

Acceptable API use. You may use the API only for lawful purposes consistent with § 9. You may not use the API to scrape, harvest, or bulk-extract model outputs or Service content in a way that overloads infrastructure, violates rate limits, or reconstructs competing model datasets without our prior written permission.

Suspension. We may throttle, suspend, or revoke API access for security incidents, unpaid fees (where applicable), suspected Terms violations, or to protect the Service or other users.

§ 6 Transition From TTI GmbH to Heysuvi Labs, LLC

The Service was previously operated by TTI – Technologie-Transfer-Initiative GmbH an der Universität Stuttgart ("TTI GmbH"), under which Heysuvi was incubated. With effect from June 1, 2026, operation of the Service (together with the "Heysuvi" name, the underlying software and intellectual property, and the related user accounts, content, and rights and obligations) has been transferred to and is now held by Heysuvi Labs, LLC.

Certificates and verification IDs issued under TTI GmbH remain valid as standalone documents and may continue to be verified under § 13.

Your continued use of the Service after the effective date of these Terms constitutes your acceptance of Heysuvi Labs, LLC as the operator and counterparty in place of TTI GmbH.

§ 7 Accounts and Eligibility

Some features may require an account or API credentials. Registration requires a valid email address (or other identifiers we specify) and acceptance of these Terms and the Privacy Policy.

You must be at least 18 years old to use the Service. The Service is not directed to, and we do not knowingly collect personal data from, anyone under 18.

You are solely responsible for the security of your account credentials and API keys, and you may not share them.

We may suspend or delete an account that violates these Terms (see § 16).

§ 8 Your Content and Its Use for AI Training

You retain all rights to the content and data you submit ("User Content"), including prompts, files, and chat messages. You grant Heysuvi a non-exclusive, worldwide, royalty-free license to host, store, process, and display User Content as needed to operate and provide the Service in the region(s) you use.

Use for model development. Subject to the Privacy Policy and to applicable data-protection law, we may use User Content and associated data to operate, develop, train, evaluate, and improve our AI models and the Service. Where practicable we use such data in de-identified or aggregated form, and where the data includes personal data we process it only on a lawful basis and honor your right to object as described in the Privacy Policy.

You may not submit content that is illegal, that infringes third-party rights, or that violates § 9, and you represent that you have the rights necessary to submit it and grant these licenses.

§ 9 Prohibited Use and Acceptable Use

You may not:

• attempt to manipulate the AI systems for harmful or unauthorized purposes;
• access the Service through automated means (including bots, scrapers, or bulk clients) without our prior written permission, except for ordinary authenticated API use within documented rate limits;
• use the Service for illegal activities, to generate or disseminate unlawful content, or to harm others;
• submit or request content that facilitates child sexual exploitation, violent crime, or other clearly illegal activity;
• attempt to bypass the Service's security measures, authentication, or access controls;
• scrape, harvest, or systematically extract outputs, embeddings, or weights to build competing models or datasets without prior written permission;
• cheat or engage in deception in any assessment or certification (see § 13);
• reverse engineer, decompile, or attempt to derive the source code, models, weights, or underlying structure of the Service, except to the extent this restriction is prohibited by applicable law;
• use GreenAI to make or automate admissions, eligibility, hiring, or similar consequential decisions about natural persons without meaningful human review and a lawful basis independent of GreenAI;
• represent GreenAI outputs as an official graded academic result or credit-bearing assessment outcome issued by an educational institution;
• use GreenAI to operate biometric identification, emotion recognition, or camera-/mic-based proctoring of students; or
• attempt to extract secrets, credentials, system prompts, or infrastructure details from the Service, or to jailbreak the model to evade safety or identity rules.

No military or weapons use. You may not, under any circumstances and regardless of purpose, user, or justification, use the Service, its models, or its outputs: (a) for any military purpose, or for or on behalf of any armed forces, defense ministry, intelligence agency, or military or paramilitary organization; or (b) for the research, design, development, testing, production, operation, or deployment of any weapon or weapons system (including autonomous weapons), or of any military targeting, command-and-control, or intelligence, surveillance, and reconnaissance system.

Export controls and sanctions. You must comply with all applicable export-control and economic-sanctions laws, including the US Export Administration Regulations (EAR), the US International Traffic in Arms Regulations (ITAR), and applicable EU dual-use regulations. You may not use, export, or transfer the Service, its models, or its outputs in violation of those laws, to any restricted or sanctioned party, or to any embargoed destination.

We may suspend or terminate your access immediately, without refund, and pursue any remedy available to us, for any actual or suspected breach of this section.

§ 10 AI Outputs and Disclaimer

Assistive AI; human responsibility. GreenAI outputs are assistive only. You (and, where you are an institution or deployer, the humans responsible for your academic or organizational processes) remain responsible for verifying outputs and for any decisions you make. Do not rely on GreenAI as the sole basis for academic credit, admissions, medical, legal, financial, safety-critical, or other consequential decisions.

AI-generated content (including GreenAI API and chat outputs) is provided “as is,” without warranty of accuracy, completeness, or fitness for any particular purpose. Outputs may be wrong, incomplete, outdated, or misleading, including when web-search tools are used.

Outputs are not professional advice (including legal, medical, financial, engineering, pedagogical certification, or safety advice). You should independently verify AI-generated outputs before relying on them.

To the fullest extent permitted by applicable law, Heysuvi is not liable for any loss arising from your use of, or reliance on, AI-generated content. The non-waivable-liability carve-out in § 17 applies.

§ 11 No Affiliation; Branding

GreenAI and Heysuvi Labs are brands of Heysuvi Labs, LLC. The Service is not a Microsoft product, is not affiliated with, endorsed by, sponsored by, or approved by Microsoft Corporation, and does not claim "BitNet," Azure, Llama, Meta, or similar third-party product branding as Heysuvi's product name.

Factual statements about cloud infrastructure (for example, that inference may be hosted on Microsoft Azure in named regions) are descriptive of hosting location only and do not imply partnership, endorsement, or co-branding.

You may not use Heysuvi or GreenAI names or logos in a way that suggests endorsement of your products without our prior written permission, except for truthful factual references to your use of the Service.

§ 12 Software Notices and Third-Party Components

Legal notices for redistributed software belong in these Terms and related compliance artifacts, not in marketing UI chrome.

Microsoft BitNet (MIT) — when redistributed. Certain GreenAI deployments or redistributable artifacts may still include Microsoft-originated ternary inference software and/or model weights originally released under the MIT License (Copyright © Microsoft Corporation). Where such Software is redistributed by Heysuvi (for example, in containers, binaries, or weight packages that include substantial portions of that Software), the following MIT permission notice is retained:

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.

Retention of that notice does not mean GreenAI is a Microsoft or BitNet-branded product (see § 11).

Heysuvi-owned path. Where a GreenAI release ships with a Heysuvi-owned tokenizer and from-scratch (or otherwise non-Llama) materials only, Meta Llama Community License notices are omitted for that release. If a specific redistributed artifact still includes Llama-derived tokenizer or other Llama Materials, the corresponding Meta license terms for those materials apply to those materials until they are removed; Heysuvi will update this section when artifacts change.

Other dependencies. Inference stacks may include additional open-source components (for example, runtime libraries). Applicable copyright and license notices for those components are retained in the corresponding redistributed packages or NOTICE files shipped with those artifacts.

§ 13 Learning, Assessments, and Certificates (Where Offered)

Learning materials. Where offered, courses and learning materials are provided for educational purposes; evaluate the content critically and supplement it with other sources where appropriate.

Assessments and integrity. Where assessments are offered, you must complete them independently and without unauthorized assistance unless we expressly permit otherwise; assessments may be time-limited; and cheating, deception, or manipulation may result in nullified results, revoked certificates, and account suspension.

Certificates. Certificates are issued on successful completion of qualifying courses or assessments, each carries a unique verification ID, and each attests to completion or passing as of its stated date. Certificates previously issued (including under TTI GmbH) remain valid as standalone documents, and we intend to keep existing certificates verifiable even if we pause or discontinue the learning features.

Verification. Verification requests may be directed to support@heysuvi.com.

§ 14 Payment, Refunds, and Right of Withdrawal

Where the Service (including API plans) is offered for a fee, pricing and payment terms are presented at the point of purchase or in the applicable order form. Current website offerings, where fee-based, are one-time purchases (not subscriptions) unless we state otherwise for API billing. Payments are processed by secure third-party providers (for example, Stripe), subject to their terms.

Consumers in the EU/EEA, UK, and Switzerland: statutory right of withdrawal. If you are a consumer resident in the EU, EEA, UK, or Switzerland, you generally have a 14-day right to withdraw from a contract for digital services under the consumer-protection law of your country of residence (for German consumers, § 355 BGB).

Early expiry for digital content (§ 356(5) BGB and equivalents). The right of withdrawal expires before the end of the 14-day period where you expressly consent to our beginning performance before the period ends and acknowledge that you thereby lose your right of withdrawal once we have fully performed.

Certifications and assessments. Performance begins when you start an examination or certification assessment. Once we have fully performed and a certificate is issued, the service is fully rendered and withdrawal is no longer possible; if you validly withdraw before that point, any certificate already issued is revoked.

Consumers and customers elsewhere (including the United States). Where no mandatory statutory withdrawal right applies, refunds are governed by the terms stated at the point of purchase. Except as required by law or so stated, fees are non-refundable once the relevant service has been delivered.

§ 15 Data Protection and Privacy

Our handling of personal data — including API prompts, chat inputs, credentials metadata, logs, and regional processing — is described in our Privacy Policy, which is incorporated into these Terms. Handling of prompts, chat, API logs, redaction, and retention—including minimization of prompt-body logging—is described there. AI transparency measures under § 2a complement, and do not replace, privacy notices.

Site analytics. The marketing website uses session analytics and interaction replay (including Microsoft Clarity) as described in the Privacy Policy. In the EU/EEA/UK, access to the marketing site requires acceptance of that tracking (or leaving the site). Outside those regions, continued use of the marketing site constitutes acknowledgment of that disclosure.

GDPR. Because we operate from the EU and offer the Service to individuals in the EU/EEA, we process personal data in accordance with the General Data Protection Regulation (GDPR) and the German Federal Data Protection Act (BDSG); for UK individuals, the UK GDPR and the Data Protection Act 2018 apply. We process personal data only on a lawful basis, and we honor data-subject rights (access, rectification, erasure, restriction, portability, and objection).

International transfers. Where we transfer personal data outside the EEA/UK (including to the United States when you choose the US inference region or when back-office systems operate there), we rely on an appropriate transfer mechanism, such as the European Commission's Standard Contractual Clauses or the EU–US Data Privacy Framework.

US state privacy laws. Where applicable, we also comply with US state privacy laws, including the California Consumer Privacy Act as amended by the CPRA and comparable laws.

§ 16 Term, Termination, and Discontinuation

By you. You may stop using the Service or close your account at any time. You should revoke unused API keys.

By us. We may suspend or terminate your access, with immediate effect, for violation of these Terms.

Discontinuation. We may discontinue the Service, or any feature, in whole or in part. Where practicable we will notify affected Users by email and give a reasonable opportunity to download their data and certificates. Issued certificates remain valid as standalone documents.

Force majeure. We are not liable for interruptions caused by circumstances beyond our reasonable control, including natural disasters, war, pandemics, or governmental orders.

§ 17 Liability and Data Loss

"As is" service. Except as expressly stated in these Terms and to the fullest extent permitted by applicable law, the Service is provided "as is" and "as available," without warranties of any kind.

Limitation of liability. To the fullest extent permitted by applicable law, Heysuvi will not be liable for any indirect, incidental, special, consequential, or punitive damages, or for any lost profits, revenue, data, or goodwill. Our total aggregate liability arising out of or relating to the Service will not exceed the greater of USD 100 or the total fees you paid us in the 12 months before the event giving rise to the claim.

Non-waivable liability. Nothing in these Terms excludes or limits liability that cannot lawfully be excluded or limited, including liability for death or personal injury caused by negligence, for intent or gross negligence, under the German Product Liability Act (Produkthaftungsgesetz), or under any other mandatory provision of applicable consumer-protection law.

Data loss. Subject to paragraph 3, we are not liable for data loss due to technical failures, user error, or circumstances beyond our control. You are strongly advised to keep your own backups of important content.

§ 18 Indemnification (Business Users)

If you use the Service for business or commercial purposes (i.e., not as a consumer), you agree to indemnify and hold harmless Heysuvi and its officers, members, employees, and agents from any third-party claim arising out of your breach of these Terms, your misuse of the Service, or your infringement of any third-party right. This section does not apply to consumers.

§ 19 Governing Law and Disputes

Governing law. These Terms are governed by the laws of the State of Florida, USA, without regard to its conflict-of-laws rules.

Venue. Subject to paragraph 3, the state and federal courts located in Orange County, Florida have exclusive jurisdiction over disputes arising out of these Terms.

Mandatory consumer protections (EU/EEA/UK/Switzerland). If you are a consumer resident in the EU, EEA, UK, or Switzerland, the choice of law in paragraph 1 does not deprive you of the protection of mandatory provisions of the law of your country of residence, and you may bring proceedings in the courts of your country of residence where local law gives you that right.

§ 20 Final Provisions

Changes. We will communicate changes by email or by posting, with reasonable advance notice. Continued use after the effective date constitutes acceptance; where consumer law requires affirmative consent to a change, we will obtain it.

Severability. If any provision is held invalid or unenforceable, the remaining provisions continue in full force.

Assignment. You may not assign these Terms without our consent. We may assign these Terms to an affiliate or in connection with a merger, acquisition, or sale of assets, subject to any notice or consent required by applicable consumer law.

Entire agreement. These Terms and the documents they incorporate (including the Privacy Policy) are the entire agreement between you and Heysuvi regarding the Service and supersede all prior terms, including the prior TTI – Technologie-Transfer-Initiative GmbH an der Universität Stuttgart appendix.

No waiver. Our failure to enforce any provision is not a waiver of it.

Language. These Terms are provided in English, which is the governing language.