中文

This is the English translation. The Chinese version prevails where permitted by applicable law.

Welcome to Simplex AI. These Terms are between you and the Simplex AI team (“Simplex”, “we” or the “Platform”) and govern your use of Simplex AI services. Contact us at contact@zqwei-tech.cn or through our Contact page.

Please pay particular attention to fees, user content, risks of model outputs, personal information, service suspension and limitations of liability. These Terms take effect between us when the registration or relevant service page prominently presents them and the important provisions, and you clearly indicate acceptance. This page identifies the current version; updates do not retroactively expand your acceptance of an earlier version. Browsing public pages alone does not mean that you accept every optional feature, paid service or non-essential use of personal information.

1.1 These Terms apply to simplexai.net and related subdomains, and to consoles, Playground tools, APIs, SDKs, experimental applications and supporting documentation that expressly refer to them. Independently supplied third-party products and services governed by separate terms are not automatically included merely because they use the Simplex name or a link.

1.2 The Platform provides or presents decision, speech, real-time interaction, memory, perception and embodied-intelligence capabilities, together with articles, research, demonstrations and benchmarks. Available capabilities are those actually offered on the product pages. “Coming soon”, “research preview” and similar descriptions do not promise delivery or a particular delivery date.

1.3 Decision services include Simplex CD (SPX-CD) and clearly identified third-party models in the model catalog. An API identifier such as spx-cd identifies the corresponding service. Access through our API does not make a third-party model a Simplex-developed model. Unavailable or suspended features, including speech tools, are subject to their feature terms only after reopening and appropriate notice.

1.4 These Terms do not replace enterprise, data-processing, private-deployment or hardware contracts, open-source licenses or service-specific rules. A separate valid written agreement prevails on matters it expressly addresses, without excluding mandatory duties or individual statutory rights.

1.5 Accounts, balances, memberships, data and permissions are not automatically shared or transferred between products because they have related names, brands or links. Any such sharing or transfer requires an explanation of its scope, operators and data flows, together with necessary authorization.

2.1 Services primarily serve developers, researchers and users capable of exercising appropriate judgment. Users under 18 should use them with a guardian's knowledge and consent; children under 14 should not register. Any future service specifically for children will have separate rules and guardian-authorization arrangements.

2.2 Register with an email address you control, provide truthful, lawful and necessary information, and maintain a valid channel for important notices. Additional identity verification is required only where legally required or necessary for a particular feature. These Terms do not independently require unrelated identity documents.

2.3 When acting for an organization, obtain appropriate authority and use its account and resources within that authority. Separate enterprise arrangements govern account ownership, membership and payment responsibility where applicable.

2.4 Protect passwords, email codes and Access Tokens/API keys. Do not disclose credentials publicly, trade accounts, or share or rent credentials to circumvent access controls, rate limits or payment rules. Lawful integration into your own application is permitted, provided you manage end-user access and keep server-side keys out of browsers, clients and public repositories.

2.5 Revoke affected Tokens, change your password and contact us promptly if you detect unauthorized access or exposed credentials. Responsibility depends on fault and applicable law. Use of a valid credential alone does not make you responsible for every resulting loss.

3.1 Subject to these Terms and applicable product rules, you receive a limited, non-exclusive permission to use the services. Available models, quotas, QPS, concurrency, context and file limits are those disclosed at use or agreed by valid contract. Section 4 governs outputs. Both free and paid use must comply with applicable restrictions.

3.2 You may build applications, perform reasonable functional testing and publish evaluations within the permission granted. Accurately describe model version, test date, datasets, sample scope, parameters and relevant conditions. Do not fabricate results or make misleading comparisons. An honest unfavorable evaluation is not a breach.

3.3 Do not use the services for unlawful activity, fraud, malware distribution, unauthorized surveillance or infringement of privacy or intellectual property. Do not submit state secrets, trade secrets, credentials or restricted content that you have no authority to process.

3.4 Do not attack the Platform, steal information, bypass safeguards, maliciously occupy computing capacity, mass-register for free resources or extract non-public weights, credentials or system information without authorization. This does not categorically prohibit legally permitted security research, interoperability or uses expressly allowed by open-source licenses.

3.5 Obtain authorization for bulk distillation, model extraction or collection beyond public permissions. This does not automatically prohibit ordinary business processing of lawfully obtained outputs or independent development of similar products.

3.6 If you offer the services to your own users, fulfill applicable notice, authorization, content-safety and consumer-protection duties, limit submitted data and provide appropriate complaint and data-rights channels. Our own duties remain ours and are not transferred to you merely because you are a developer.

4.1 “User content” includes text, images, audio, video, files, context, questions, options, prompts, business data and other material you are entitled to submit. “Outputs” include returned decisions, scores, candidate distributions, reasoning text, generated material and other results.

4.2 Uploading content or using the services does not transfer your or another rightsholder's existing intellectual property or other lawful rights to us. You must have a lawful basis for submitting, processing and instructing us to process content, including required permissions for others' personal information, likeness, voice or works.

4.3 You permit the receipt, transmission, inference, format conversion and storage necessary to perform your requested function, and disclosure to notified providers of the parts necessary for their roles. Enabled content archives may store requests, answers and images, subject to the Privacy Policy. This permission does not include unrestricted resale, public display or perpetual use for any purpose.

4.4 Participation preference in our data improvement program is initially enabled. That default and acceptance of these Terms do not replace any valid authorization legally required for training, labeling or evaluation. With an applicable lawful basis and within authorized purposes, program content may be retained long term. Turning the preference off excludes existing and new content from new training use and normally limits confidential retention to 30 days from each call. Rejoining applies only to subsequent new calls. Necessity, third-party information, completed training and copies are governed by sections 3 and 5 of the Privacy Policy. Refusing optional training does not affect basic services that do not depend on it.

4.5 To the extent permitted by law, you may use, copy, modify and commercially use outputs returned to you. We do not restrict such use of output rights we lawfully hold and may dispose of, subject to these Terms. We do not guarantee that every output is protectable, unique or free of third-party rights. Additional third-party or service-specific conditions should be disclosed before selection or connection.

4.6 Software, algorithms, models, documentation, website design, marks and other technology belong to us or their lawful rightsholders. Lawfully developed general improvements that contain no identifiable personal information, original user content or others' protected information do not automatically belong to a particular user because they reflect general service issues. Calling information “derived data” does not confer unlimited rights to user content or personal information.

4.7 General feedback you voluntarily submit may inform service improvements. Personal information, non-public samples, confidential business material and protected works do not gain an additional license merely by being called feedback. Applicable licenses govern articles, demonstrations, code and datasets; public accessibility alone is not a permission to copy without restriction.

5.1 AI outputs may be inaccurate, incomplete, biased or unstable. Probabilities, scores and confidence from Simplex CD (SPX-CD) are calculations under particular model and candidate conditions, not guaranteed real-world accuracy or assurances about consequences. Validate results for your task; a high score alone is insufficient for consequential action.

5.2 Outputs do not replace qualified medical, legal or financial advice. For employment, education, credit, insurance, healthcare, justice or other decisions with major effects on individuals, establish a lawful basis, appropriate human review and routes for explanation, appeal or correction.

5.3 Use permissions, safeguards, human confirmation, emergency stops and fallback mechanisms proportionate to risk when outputs control robots, vehicles, machinery, payments, deletion or other real-world actions. Unless expressly agreed in writing and appropriately certified, a general model API is not a certified safety controller and must not be the sole decision or control mechanism for actions that could endanger people.

5.4 Do not use synthesized voices or digital likenesses to impersonate without lawful authorization, defraud, mislead or infringe personality rights. Inferences about voice, emotion, gaze or behavior are limited and must not be treated as definitive statements of a person's mental state, health or identity.

5.5 Each party must fulfill applicable generated-content labeling duties for its service and role. Before opening a feature, the Platform should implement its required notices and labels; these Terms do not claim that all outputs already have visible or embedded labels. Fulfill applicable disclosure duties when distributing content publicly, and do not maliciously remove, alter, forge or conceal legally required labels.

5.6 If a service lawfully allows delivery without visible labels, a specific feature and agreement must define your labeling duties and responsibilities, with records retained as required. These Terms do not authorize you to remove legally required labels yourself.

6.1 Free trials, grants, discounts and testing access apply only within their disclosed scope, duration and conditions; they are not promises of permanently free access. Reasonable QPS, concurrency, security and resource limits may still apply.

6.2 During public beta only, each account shares 50,000,000 free input tokens per day across SPX-CD Flash and Pro API and Playground usage, resetting at 00:00 Beijing time. Overage input costs USD 0.025 per million tokens for Flash and USD 0.08 for Pro; output tokens are free. Auto routing uses the actual model price. After the daily free allowance, overages automatically charge USD balance, including invitation credits, at the disclosed prices. Overage calls stop without sufficient available balance. Confirmed actual input is billed; failed requests are free. Uncertain execution or usage retains reserved funds pending reconciliation. Fractional-cent costs accumulate and whole cents are deducted, with the remainder carried forward. Top-ups add available balance. Historical free calls made before this billing mechanism took effect are not charged retroactively. Prices and allowances after public beta will be disclosed and confirmed separately.

6.3 Before charging begins, we should clearly disclose the service, units, prices, taxes, calculation basis, currencies, credit expiry and deduction rules. Conversion between USD balances and CNY prices must be explained before payment or charging authorization. An announced future price or existing balance alone does not authorize charges without necessary notice and confirmation.

6.4 Usage-based services you confirm may deduct balance under disclosed rules. Whether input, output, cache or retries are charged must be explained. Treatment of failed or partially successful requests must not rely on undisclosed internal rules. You may request investigation, correction and refunds for duplicate or erroneous charges or anomalous usage records.

6.5 Subscriptions, automatic top-ups and recurring payment-tool deductions require separate clear authorization, including amounts or calculation methods, frequency and cancellation, with legally required reminders. Silence does not mean acceptance of automatic renewal. Cancellation stops new renewal charges without extinguishing lawfully incurred amounts.

6.6 Paid credit is generally for permitted Platform services, earns no interest and must not facilitate unlawful transfers. You may request a refund of unused paid credit; we will verify identity, source of funds, consumed services and applicable refund obligations. Grants and unpaid coupons are not directly redeemable for cash, but this does not reduce refundable amounts you actually paid.

6.7 Completed computing or consumed services are not generally refundable solely because of subjective preference. This does not limit refunds for duplicate or unauthorized charges, failure to provide agreed services, refundable prepayments or other statutory grounds. Exceptions for customized, immediately performed or digital services apply only when lawful and appropriately disclosed and confirmed, not as blanket exclusions.

6.8 If we stop a service and cannot fulfill purchased rights, we will address unperformed services and refundable balances as required. Force majeure, termination or account closure does not automatically forfeit credit. Refunds normally return through the original payment channel; difficulties should be resolved through another lawful method with you. Provider or bank delays and lawful fees should be explained, without undisclosed punitive charges.

7.1 We handle personal information under the Simplex AI Privacy Policy. Accepting these Terms does not replace legally required separate consent, guardian consent or other necessary authorization.

7.2 Some features use third-party infrastructure, sign-in, payment or model providers. Before choosing a third-party model, consider its identity, data recipients, processing locations and restrictions. Jev is provided by TypeSafe AI: necessary request content is sent directly to TypeSafe AI or through Vercel AI Gateway to TypeSafe AI. Vercel processes a request only when that request uses the intermediary route. The actual connection and feature disclosures determine the route, subject to the Privacy Policy and feature-specific notices.

7.3 We should provide necessary notice and accessible information for independent third-party rules. Upstream outages or rule changes may affect availability; we will respond reasonably without using “third-party causes” to automatically disclaim our own selection, management, notice or performance duties.

7.4 For end-user information submitted by enterprises, responsibilities depend on who actually determines purposes and means. A data-processing agreement should be concluded where needed. Describing a relationship as entrusted, independent or joint processing does not by itself change legal responsibility.

8.1 Research previews and test features may involve model changes, performance changes, interface changes and temporary unavailability. Without a separate written agreement, we do not promise fixed accuracy, latency or production service levels, while retaining reasonable service, security and legal duties.

8.2 We will give reasonable advance notice and appropriate arrangements for major changes affecting purchased rights, subject to stricter legal requirements. Free features may reasonably change or end, without misleading promotion or retroactive charges.

8.3 To address security events, unlawful use, abnormal traffic, maintenance or lawful authority requests, we may impose proportionate limits, suspend Tokens or pause features. We should notify in advance where feasible, or promptly explain emergency measures afterward, without unnecessarily affecting unrelated services and data.

8.4 You may stop using the service and request account closure. Before closure, we will identify outstanding amounts, refundable credit and data that need export. Required records may remain only for their applicable purposes and periods.

8.5 You may appeal through our contact channels. We should investigate and explain our action, and promptly correct restrictions imposed in error. Termination does not extinguish prior lawful rights and duties or provisions intended to survive, including confidentiality, intellectual property and dispute handling.

9.1 Each party is responsible for breach according to applicable law and fault. Claims should establish breach, loss and causation, reasonably mitigate loss and avoid double recovery.

9.2 We take reasonable measures proportionate to service risk to protect services and data, without guaranteeing fault-free internet or model services. Force majeure, notice, mitigation and responsibility follow applicable law; technical failures, vendor issues and business changes are not automatically force majeure.

9.3 Any general contractual liability cap for paid enterprise services should be expressly negotiated in a separate valid agreement with appropriate notice. Free services also remain subject to non-excludable duties. No agreement excludes or limits responsibility that cannot lawfully be excluded or limited, or offsets refundable unused payments and other mandatory obligations.

9.4 These Terms do not remove our applicable cybersecurity, personal-information, content-governance or service duties, or your rights to complain, litigate and seek other remedies.

10.1 We may update these Terms for product or legal changes, identify update and effective dates, and appropriately notify important changes. New charges, material reductions of purchased rights and matters requiring renewed agreement need appropriate confirmation. Prior acceptance is not automatic authorization for new personal-information purposes.

10.2 If you disagree with important changes, you may stop the relevant service and seek lawful handling of unfulfilled purchases and data. Earlier disputes are generally governed by the applicable earlier terms, without adverse retroactive application.

10.3 These Terms are governed by the laws of mainland China. We should first seek an amicable resolution of disputes, then use a People's Court with lawful jurisdiction if unresolved. Mandatory local protections that cannot be excluded by contract remain applicable to overseas users where required.

10.4 Where language versions differ, the Chinese version prevails to the extent permitted by law. Translation differences must not mislead users or reduce statutory rights. Invalidity of one provision does not invalidate independently effective provisions. Interpretation follows law; the Platform has no unilateral “final interpretation” right.

10.5 Send questions, complaints, refund requests, infringement notices and other requests to contact@zqwei-tech.cn or our Contact page. Include necessary account, order or request identifiers, never passwords, verification codes or full API keys. Personal-information requests follow the procedures and time requirements of the Privacy Policy.

Read the Privacy Policy ↗Back to top ↑