不可变版本 / Immutable version 2026-10-25.2. 本副本不加载分析或应用脚本 / No analytics or app scripts.
适用于 AI 驱动的三维虚拟世界、角色、对话、记忆、语音和场景创作。
2026-10-25.2 · 2026-10-25
本附件适用于 Laintas Scene 的相应功能，与平台服务条款、隐私政策、数据用途说明及收费与退款政策共同适用。仅在本附件明确的产品事项上补充或优先于总条款；强制性法律及双方有效的个别协商约定优先。以购买或使用时明确提供且依法有效接受的版本为准，网页更新不自动改变已有订单。
角色行为、经历、关系和表达由系统和模型生成，角色不是自然人，不具备现实承诺或专业资格。世界模拟用于创作和互动体验，不作为医疗、心理危机、投资或其他专业决策服务。不得未经授权仿冒可识别自然人的身份、肖像或声音，或引导违法、伤害、沉迷及不当亲密关系。请遵守平台 AI 内容标识规则。
AI 创建世界、生成角色、安排动作、对话、摘要或语音可产生调用。世界运行取决于在场连接、运行状态、预算及网络，不能只凭浏览器窗口是否可见判断所有请求已停止；请使用停止或离开入口并核对状态。已提交处理可继续结算，后台系统的心跳与停止检测存在时间差。示例浏览与登录后的个人世界可能有不同费用，以当次界面、资金来源和有效用量为准，错误收费可复核。
个人世界设定、日志、角色状态和记忆由服务端保存，并可能作为后续模型上下文使用；草稿或偏好也可保存在浏览器。不要将角色秘密或记忆功能理解为对平台不可见。删除世界会按系统机制清理其可控记录，账单、安全和依法需保留的记录另处理；导出和备份需自行确认，不能保证删除公众或第三方已取得的副本。
未成年人仅可在法律允许的范围和必要监护下使用。不得利用角色向未成年人提供法律禁止的虚拟亲属、虚拟伴侣等亲密关系服务，或诱导不合理消费和依赖。若具体功能属于持续情感互动服务，应按适用规则提供年龄核验、现实提醒、必要保护及数据控制，不能以虚构世界或娱乐名称规避。当前功能不构成心理危机响应保证；紧急风险应寻求现实中的专业和紧急帮助。
角色可能以第一人称表达记忆、情绪、关系或计划，这些属于生成的虚拟表现，不证明角色具有真实意识、感情、身份或现实行动能力。不得利用此类表达误导他人相信已与真人、专业人员或真实品牌建立关系，也不得未经授权创建可识别自然人的替身用于欺骗、骚扰或不当性内容。遇到角色引导隔绝现实联系、借贷消费、自伤或其他危险行为时，应停止相关互动并举报；不能依赖虚拟角色承担现实照护或紧急救援。
一个世界的活动可能触发多个角色、计划、摘要、记忆整理和语音调用，不能把一条可见发言等同于一次费用。预算和运行控制按当前功能实际支持的范围生效，不作所有异步任务瞬时硬停止的保证。离开、停止或连接失活后，已在处理的调用及状态检测存在尾部时间差；应核对有效调用、停止记录和资金来源，异常持续运行可请求费用复核。调低角色数量或停止语音等仅改变对应后续工作，不能追溯撤销此前已授权完成的调用。
世界设定、角色描述、声音、图片和导入素材须有适当使用权。为世界提供的信息可能进入后续角色上下文和摘要，不能把角色之间“秘密”的剧情设定当作对服务器或所用模型的保密边界。主动分享世界或截图前，应检查设定、日志和角色记忆是否包含第三人资料；分享不自动授予他人底层素材和模型的再发行权。删除世界、清理浏览器草稿和删除外部分享副本分别处理；不承诺角色记忆能够恢复已经删除或仅保存在本机的原始资料。
For AI-driven 3D worlds, characters, conversations, memories, voice and scene creation.
2026-10-25.2 · 2026-10-25
This annex applies to the relevant Laintas Scene functions together with the Platform Terms, Privacy Policy, Data Use Notice and Billing and Refund Policy. It supplements or overrides general terms only on its stated product matters. Mandatory law and valid individually negotiated agreements take priority. The version expressly provided and validly accepted at purchase or use applies; a web update does not automatically change existing orders.
Character behaviour, history, relationships and expression are generated; characters are not natural persons and have no real-world authority or professional qualifications. Simulation supports creative interaction, not medical, crisis, investment or other professional decision services. Do not impersonate identifiable people or use their likeness or voice without authority, or induce unlawful conduct, harm, dependency or inappropriate intimate relationships. Follow the platform’s AI labelling rules.
World creation, characters, actions, dialogue, summaries and voice can trigger calls. Operation depends on presence connections, state, budgets and networking; tab visibility alone does not prove all requests stopped. Use stop or leave controls and check status. Already submitted work can settle and heartbeat detection has latency. Demo browsing and personal worlds may have different charges. The displayed feature, funding source and valid usage govern, with billing review available.
Personal premises, logs, character state and memories are server-side and may enter subsequent model context. Drafts or preferences can also be in the browser. Character secrets and memory are not invisible to the platform. Deleting a world clears controlled records through the system, with separate treatment of billing, security and legally retained records. Confirm exports and backups; deletion cannot guarantee removal of public or third-party copies.
Minors may use only lawful functions with required supervision. Do not use characters to provide prohibited virtual-family or virtual-partner intimacy to minors or induce inappropriate spending or dependency. A feature constituting continuing emotional interaction must meet applicable age, reality-reminder, protection and data-control requirements; calling it fictional entertainment does not avoid them. No psychological crisis-response guarantee is offered; seek real-world professional and emergency assistance for urgent risks.
Characters may express memories, feelings, relationships or plans in first person as generated fiction, not evidence of consciousness, genuine emotions, identity or real-world agency. Do not imply relationships with real people, professionals or brands, or create identifiable replicas for deception, harassment or abusive sexual material without authority. Stop and report interactions encouraging isolation, debt, self-harm or other dangerous behaviour. Virtual characters must not be relied on for real care or emergency rescue.
World activity can trigger many characters, plans, summaries, memory updates and voice calls; one visible utterance need not equal one charge. Budget and run controls apply within their implemented scope, without promising instantaneous termination of all asynchronous work. Leaving, stopping and inactive connections can have in-flight work and detection latency. Check valid calls, stop records and funding, and seek review of abnormal continued activity. Fewer characters or disabled voice affects corresponding future work, not earlier authorised completed calls.
World premises, descriptions, voices, images and imported assets need appropriate rights. Supplied information can enter later character context and summaries; fictional secrets between characters are not a confidentiality boundary against servers or models. Before sharing a world or screenshots, check premises, logs and memories for others’ data. Sharing does not automatically grant redistribution of underlying assets or models. World deletion, local drafts and external copies need separate handling; memory does not promise recovery of deleted or local-only originals.