Terms of Service

    Effective June 19, 2026

    These Terms of Service (the "Terms") govern access to and use of the services, APIs, software, and websites (collectively, the "Services") provided by Lilac Research Inc. ("Lilac," "we," "us"). By creating an account, accessing, or using the Services, you ("Customer," "you") agree to these Terms.

    If you use the Services on behalf of an organization, you represent that you have authority to bind that organization, and "you" refers to that organization.

    1. The Services

    Lilac provides a platform for running AI workloads on compute capacity operated by Lilac and by third-party infrastructure and GPU capacity providers ("Capacity Providers"). The Services include, and may in the future include, a real-time, OpenAI-compatible inference API, asynchronous and batch processing, customer-defined container and training jobs, model fine-tuning, and associated storage and orchestration.

    Certain Services are subject to additional terms, pricing, or data-handling descriptions presented in the Services or in our documentation ("Service-Specific Terms"). Service-Specific Terms are part of these Terms and control if they conflict with these Terms for that Service. We may change the models, endpoints, features, and Capacity Providers at any time.


    2. Eligibility & accounts

    You must be at least 18 years old and able to form a binding contract. You must provide accurate account information and keep your credentials and API keys secure. You are responsible for all activity under your account, and you must promptly notify us at security@getlilac.com of any suspected unauthorized use.


    3. Fees, credits & payment

    Usage is billed per the pricing published in the Services (for example, per-token pricing for inference; per-GPU-time or per-job pricing for batch and container workloads). Prepaid credits, where offered, are non-refundable except where required by law and expire as stated at purchase. You authorize our payment processor (Stripe) to charge your payment method for amounts due. Late or failed payments may result in suspension. Taxes are your responsibility, excluding taxes on Lilac's income.


    4. Your content & data handling

    Ownership. "Customer Content" means the prompts, inputs, files, datasets, container images, and code you submit, batch inputs and results, fine-tuned model artifacts created for you, and model outputs processed through the Services. As between the parties, you own your Customer Content. You grant Lilac the limited right to host and process Customer Content solely to provide, secure, and support the Services. Lilac does not use Customer Content to train models and does not sell Customer Content.

    Data handling by Service type. Retention differs by Service:

    • Real-time inference API: request and response bodies are processed transiently in memory; Lilac does not store prompts or completions at rest after the response is returned, as further described in the Privacy Policy.
    • Batch, container, training & storage Services: by their nature these Services store the Customer Content you submit and the results they produce until you retrieve or delete them. Unless Service-Specific Terms state otherwise, stored job inputs and results are retained for 30 days after job completion and then deleted, and you may delete them earlier via the Services.
    • Operational metadata (for example, token counts, job and workload identifiers, resource usage, latency, and error codes) is retained for billing, reliability, security, and abuse prevention.

    Capacity Providers. Workloads may execute on Capacity Provider infrastructure. Capacity Providers are contractually prohibited from retaining, inspecting, or using Customer Content, and process it only transiently to execute your workloads.


    5. Your workloads (batch, container & training)

    For Services that execute workloads you define (batch jobs, container jobs, fine-tuning or training runs):

    • You are responsible for what you submit. You represent that you have all rights to the code, container images, datasets, model weights, and other materials you submit, and that they do not contain malicious code and will not be used to violate the Acceptable Use section.
    • Scheduling and preemption. Workloads may be queued, scheduled across varying hardware and Capacity Providers, paused, preempted, retried, or relocated at any time, including mid-run. Lilac does not guarantee start times, completion times, or uninterrupted execution unless a Service-Specific Term says otherwise. You are responsible for designing workloads to tolerate interruption (for example, checkpointing and idempotent runs).
    • Resource limits. Lilac may set and change quotas, runtime limits, storage limits, and concurrency limits, and may terminate workloads that exceed them or degrade the platform or other customers.
    • Inspection for safety. Lilac may scan submitted images and job specifications for malware and policy violations and may refuse or terminate workloads that fail such checks. Lilac does not access the contents of your datasets or job payloads except as needed to operate, secure, and support the Services or as required by law.
    • Fine-tuned artifacts. Model artifacts produced by your fine-tuning or training jobs from your data are your Customer Content, subject to the license terms of any base model used. You are responsible for complying with base model licenses.

    6. Acceptable use

    You will not, and will not permit others to:

    • (a) violate applicable law or third-party rights;
    • (b) generate or distribute child sexual abuse material or content that sexualizes minors;
    • (c) use the Services to develop, plan, or carry out violence, weapons development, or other serious harm;
    • (d) attempt to gain unauthorized access to the Services, other customers' data or workloads, or Capacity Provider environments;
    • (e) probe, scan, or test the vulnerability of the Services without written authorization;
    • (f) interfere with or disrupt the Services, including submitting workloads designed to escape isolation, mining cryptocurrency without our written approval, or consuming resources beyond purchased quotas;
    • (g) misrepresent AI-generated output as human-generated where this would deceive or harm;
    • (h) resell bare access to the Services or raw compute capacity without our written consent (building and selling your own products and applications on the Services is permitted); or
    • (i) use output or artifacts to develop competing models in violation of applicable model licenses.

    You are responsible for your end users' compliance.


    7. Model output & third-party models

    AI-generated output is probabilistic and may be inaccurate, incomplete, or offensive. You are responsible for evaluating output for your use case, including human review where decisions affect individuals. Output is provided without warranty. As between you and Lilac, and to the extent permitted by law, Lilac assigns you its right, title, and interest in output, subject to third-party model licenses.

    The Services run on infrastructure operated by Lilac and by Capacity Providers, and serve models subject to their own licenses (for example, open-weight model licenses). Capacity and models may be added, drained, or withdrawn at any time; Lilac does not guarantee any particular Capacity Provider, region, or hardware.


    8. Availability, support & suspension

    The Services are provided without an uptime commitment unless a separate written SLA is agreed; current status is published at Lilac's status page. Support is provided via support@getlilac.com on a commercially reasonable basis. Features identified as alpha, beta, preview, or early access are provided as-is, may change or be discontinued without notice, and may have different or no data-handling commitments, as stated in the applicable Service-Specific Terms.

    We may suspend or limit the Services or specific workloads immediately if we reasonably believe that your use violates the Acceptable Use section or threatens the security or integrity of the Services, other customers, or Capacity Providers; that amounts due are unpaid; or that suspension is required by law. We will use reasonable efforts to notify you and restore service once the issue is resolved.


    9. Intellectual property, feedback & confidentiality

    Lilac retains all rights in the Services. You receive a limited, non-exclusive, non-transferable right to use the Services per these Terms. If you provide feedback, we may use it without restriction or obligation.

    Each party will protect the other's non-public information with reasonable care and use it only as needed to perform under these Terms. This does not limit Lilac's data commitments in the "Your content & data handling" section.


    10. Privacy

    Processing of personal data is described in the Privacy Policy. Where Lilac processes personal data in Customer Content on your behalf, the parties may execute Lilac's Data Processing Addendum.


    11. Disclaimers & limitation of liability

    The Services, all workload execution, and all output are provided "as is" and "as available." Lilac disclaims all warranties, express or implied, including merchantability, fitness for a particular purpose, non-infringement, and any warranty arising from course of dealing, including that the Services will be uninterrupted or error-free, that workloads will start or complete by any time, or that output will be accurate or reliable.

    To the maximum extent permitted by law: (a) neither party is liable for indirect, incidental, special, consequential, exemplary, or punitive damages, or lost profits, revenue, data (including lost checkpoints or interrupted workloads), or goodwill; and (b) each party's total liability under these Terms is limited to the amounts you paid Lilac for the Services in the twelve (12) months before the event giving rise to liability (or US $100 if greater). These limits do not apply to your payment obligations, your breach of the Acceptable Use section, or either party's gross negligence or willful misconduct.


    12. Indemnification

    You will defend and indemnify Lilac against third-party claims arising from your Customer Content; the workloads, code, images, datasets, and model weights you submit; your products; or your breach of the Acceptable Use section.


    13. Term, termination & data deletion

    These Terms apply while you use the Services. Either party may terminate at any time on notice (you may simply stop using the Services and close your account). On termination, unpaid amounts become due, running workloads may be stopped, and stored Customer Content (for example, batch inputs and results, datasets, and fine-tuned artifacts) is deleted within 30 days, except minimal records retained for legal, billing, or security purposes. Sections that by their nature should survive — including Fees, content ownership, Acceptable Use, Model output, IP, Disclaimers, Limitation of liability, Indemnification, and these General terms — survive termination.


    14. Changes to these Terms

    We may update these Terms or Service-Specific Terms by posting a revised version with a new effective date; material changes will be notified via the Services or email. Continued use after the effective date constitutes acceptance.


    15. General

    These Terms are governed by the laws of the State of Delaware, excluding conflict-of-laws rules, and the parties consent to exclusive jurisdiction and venue in the state and federal courts located in Wilmington, Delaware. You may not assign these Terms without our consent except to a successor in a merger or asset sale; we may assign to an affiliate or successor. These Terms (with the Privacy Policy, any Service-Specific Terms, and any order or Data Processing Addendum) are the entire agreement. Export control and sanctions laws apply to your use. If any provision is unenforceable, the remainder stays in effect.


    16. Contact

    Mailing address

    Lilac Research Inc.
    2261 Market Street STE 60372
    San Francisco, CA 94114

    A Delaware C Corporation.