User Agreement / 用户协议

Overview / 概览

We follow a data-minimization approach. We do not sell personal data, do not run ads, and do not build marketing profiles. We process data only as needed to provide the service, keep it secure, and prevent abuse.

我们遵循最小必要原则处理数据。我们不会出售个人数据，不投放广告，也不会建立营销画像。我们仅在提供服务、保障安全和防止滥用所必需的范围内处理数据。

Important Notice / 重要提示

• By tapping "Agree to Terms and Enter", you confirm that you have read, understood, and expressly agreed to this agreement. Agreeing to this agreement allows you to enter the app. Authorization for third-party AI and search processing may be granted on the agreement page or, if not granted there, requested again before your first message is sent using those services.
• To provide AI generation, web search, workflow, and related features, your inputs, submitted content, and the context reasonably necessary to generate a reply may be sent to OpenRouter, relevant underlying model providers, Jina AI, and other necessary third-party service providers. If you do not grant third-party processing authorization on the agreement page, the app will ask again before sending a message that requires such processing.
• Group chats may allow you to upload a single image. Uploaded images may be safety-checked by our server and analyzed by a vision or multimodal model to generate a short caption or OCR summary used as message context.
• This app is published and intended only for users who are 18 years of age or older.
• Do not enter sensitive personal information, and do not enter copyrighted, trademarked, or otherwise protected content that you are not authorized to use.

• 点击“同意协议并进入”，即表示您已阅读、理解并明确同意本协议。完成协议同意后，您可以进入应用；对于第三方 AI / 搜索处理授权，您可以在本页一并同意，或在首次发送消息前再次确认。
• 为实现 AI 生成、联网搜索、工作流及相关功能，您输入的内容、您主动提交的内容，以及生成回复所合理必需的上下文信息，可能被发送至 OpenRouter、相关底层模型提供商、Jina AI 及其他必要的第三方服务提供商。如果您未在本页同意第三方处理授权，应用会在首次发送需要相关处理的消息前再次征求您的确认。
• 群聊功能可能允许您上传单张图片。已上传的图片可能会先经过我们服务端的内容安全审核，并由视觉或多模态模型生成简短 caption 或 OCR 摘要，作为消息上下文的一部分参与处理。
• 本应用仅面向年满 18 周岁的用户发布和提供服务。
• 请勿输入敏感个人信息，也请勿输入任何您无权使用的受版权、商标或其他知识产权保护的内容。

If you do not agree, do not use the app and please close it.

如果您不同意本协议，请不要使用本应用，并请立即关闭。

1. Cloud Storage and Sync / 云端存储与同步

To enable multi-device access, session recovery, chat history, workflow execution, and service reliability, conversation content and related metadata are stored on our servers. If you upload an image in a group chat, the image itself, related image metadata, and the derived short caption or OCR summary may also be stored as part of the conversation record. Please do not include sensitive personal information in your messages or uploaded images, such as government identifiers, passwords, bank card details, health data, biometric data, or other highly sensitive data.

Where reasonably necessary for service operation, security logging, abuse handling, or compliance, stored conversation data, prompts, metadata, and operational records may be retained in redacted, masked, tokenized, aggregated, or otherwise de-identified form. We aim to keep only the minimum amount of identifiable information necessary to provide the service, maintain account and session integrity, investigate abuse, and comply with applicable law.

为实现多设备访问、会话恢复、聊天记录保存、工作流执行和服务稳定性，您的对话内容及相关元数据会存储在我们的服务器上。如果您在群聊中上传图片，图片本身、相关图片元数据以及派生的简短 caption 或 OCR 摘要，也可能作为会话记录的一部分被存储。请不要在消息或上传图片中包含敏感个人信息，例如身份证件号码、密码、银行卡信息、健康数据、生物识别信息或其他高度敏感的数据。

在为实现服务运行、安全日志、滥用处置或合规义务所合理必需的范围内，存储的对话数据、提示词、元数据和运行记录，可能以脱敏、遮蔽、标记化、聚合或其他去标识化形式保存。我们会尽量仅保留提供服务、维持账号和会话完整性、调查滥用行为以及遵守适用法律所必需的最少可识别信息。

2. Third-Party AI Processing and Explicit Authorization / 第三方 AI 处理与明确授权

Third-party AI and search processing requires your explicit authorization. You may grant that authorization on the agreement page. If you do not grant it there, the app will ask again before the first message that requires such processing is sent.

To provide AI generation features, we may send your inputs, submitted content, prompts, conversation history, summaries, reply metadata, workflow context, tool parameters, uploaded images, image-derived short captions or OCR summaries, and other context reasonably necessary to generate a response to OpenRouter and relevant underlying model providers. We may also use technical routing, monitoring, or observability channels when accessing those services.

The live version of the service may involve underlying model providers including OpenAI, Google, DeepSeek, xAI, Alibaba Cloud, and Moonshot AI. Their handling of data may be subject to their own terms, privacy policies, and operational practices.

Data processed by us or stored on our servers is not used to train general-purpose AI models. For third-party AI processing used in the live service, we intend to use service configurations, contractual terms, or provider settings that prohibit use of our user data for model training or model improvement. If a processing path does not meet that requirement, we will not intentionally use it for user content in the live service until an equivalent no-training arrangement is available.

第三方 AI / 搜索处理需要您的明确授权。您可以在协议页直接授予该授权；如果您未在该页面授予，应用会在首次发送需要相关处理的消息前再次征求您的确认。

为提供 AI 生成能力，我们可能会将您的输入内容、主动提交的内容、提示词、对话历史、摘要、回复元数据、工作流上下文、工具参数、上传的图片、图片派生的简短 caption 或 OCR 摘要，以及生成回复所合理必需的其他上下文信息发送至 OpenRouter 及相关底层模型提供商。在访问这些服务时，我们也可能使用技术路由、监控或可观测性通道。

当前上线版本涉及的底层模型提供商可能包括 OpenAI、Google、DeepSeek、xAI、Alibaba Cloud 和 Moonshot AI。相关第三方如何处理数据，可能受其各自的服务条款、隐私政策和实际运营规则约束。

我们处理或存储在自有服务器上的数据，不会被用于训练通用人工智能模型。对于上线版本接入的第三方 AI 处理服务，我们将以禁止将用户数据用于模型训练或模型优化的服务配置、合同条款或供应商设置为准入前提；如果某条处理链路不满足该要求，我们不会有意将用户内容发送至该链路，直至具备等效的“禁止训练”保障为止。

3. Network Search and External Tools / 联网搜索与外部工具

When the service uses web search, page reading, or other external tools, related queries, URLs, and necessary request parameters may be sent to Jina AI and other required third-party service providers so the feature can work. Results returned by those services may be included in model context and may also be saved in conversation records.

当服务使用联网搜索、网页读取或其他外部工具时，相关查询词、URL 及必要的请求参数可能会被发送至 Jina AI 及其他必需的第三方服务提供商，以实现相应功能。这些服务返回的结果可能会被纳入模型上下文，也可能被保存到对话记录中。

4. Account and Device Identifiers / 账号与设备标识

Account identifiers are used only to distinguish accounts and maintain sessions. If guest device binding is enabled, we may store an HMAC-derived hash of your device identifier and your device type to keep guest access consistent across sessions. We do not store the original raw device identifier for this purpose.

账号标识仅用于区分账户和维持会话。如果启用了游客设备绑定能力，我们可能会存储由设备标识经过 HMAC 处理得到的摘要值以及设备类型，以便在不同会话中保持游客访问的一致性。为此目的，我们不会存储设备原始标识本身。

5. Safety and Abuse Prevention / 内容安全与滥用防护

We may apply basic automated content safety filtering and anti-abuse measures to user inputs, uploaded images, and AI outputs to reduce obviously unlawful, harmful, or inappropriate content. Uploaded images may be checked before they are accepted into the conversation flow. These measures do not guarantee complete detection, blocking, or correction. We may also perform necessary security checks and logging to protect the service, other users, and our systems.

We maintain a zero-tolerance policy toward illegal, exploitative, abusive, or otherwise prohibited content and conduct. This includes, without limitation, prohibited prompts, generated outputs, uploaded material, shared character profiles, role setup text, display names, descriptions, shared preview pages, and repeated or organized abusive behavior. To enforce this policy, we may use automated detection, user reporting, manual review, and periodic backend patrols or audits to inspect content, shared character profiles, related shared artifacts, and related accounts or devices. If we reasonably believe content or behavior violates this agreement, applicable law, or platform rules, we may immediately remove content, disable sharing, revoke share codes, restrict generation, freeze features, suspend or permanently ban related character profiles, accounts, devices, or associated access, and preserve relevant records as necessary for security, investigation, or compliance.

我们可能会对用户输入、上传图片和 AI 输出应用基础的自动化内容安全过滤与滥用防护措施，以尽量减少明显违法、有害或不当内容的出现。上传图片在进入会话流之前也可能先经过安全检查。但这些措施并不保证能够完整识别、拦截或纠正所有问题内容。为保护服务、其他用户及我们的系统，我们也可能进行必要的安全检查和日志记录。

我们对违法、剥削性、滥用性或其他被禁止的内容与行为采取零容忍态度。这包括但不限于：违规提示词、生成结果、上传材料、已分享的角色内容、角色设定文本、显示名称、描述、分享预览页，以及重复性或组织化的滥用行为。为执行该政策，我们可以结合自动检测、用户举报、人工审核，以及后端定期巡查或审计机制，对相关内容、已分享的角色内容、分享产物及关联账号或设备进行检查。如果我们有合理理由认为相关内容或行为违反本协议、适用法律或平台规则，我们可立即删除内容、关闭分享、撤销分享码、限制生成、冻结功能，并对相关角色内容、账号、设备或关联访问权限采取暂停或永久封禁措施；在安全、调查或合规所必需时，我们也可以保留相关记录。

6. User Content Restrictions / 用户内容限制

You must not input, upload, generate, request, or transmit any illegal, harmful, or inappropriate content, including violent or graphic abuse, terrorism or extremism, self-harm or suicide instructions, sexually explicit content, child sexual content, hate or discrimination, harassment, fraud, phishing, malware, invasion of privacy, infringement of intellectual property rights, or other content that violates applicable law, public order, or platform rules.

Please do not input copyrighted, trademarked, or otherwise protected content that you are not authorized to use, including protected character names, brand identifiers, work titles, catchphrases, or similar material.

The restrictions in this section apply equally to character profile settings, role setup text, instructions, knowledge files, display names, descriptions, avatars, shared preview content, uploaded images, image-derived text, and any other content you create, upload, or publish through the service.

您不得输入、上传、生成、请求或传播任何违法、有害或不当内容，包括但不限于：暴力或血腥虐待、恐怖主义或极端主义、自残自杀指引、色情露骨内容、涉及未成年人的性相关内容、仇恨歧视、骚扰辱骂、诈骗、钓鱼、恶意代码、侵犯他人隐私、侵犯知识产权，或其他违反适用法律法规、公序良俗或平台规则的内容。

请勿输入受版权保护的角色、品牌或作品名称，以及其他任何您无权使用的受版权、商标或其他知识产权保护的内容。

本节限制同样适用于上传图片、角色设定、角色说明文本、指令、知识文件、显示名称、描述、头像、分享预览内容、图片派生文本，以及您通过本服务创建、上传或发布的任何其他内容。

7. AI-Generated Content Disclaimer / AI 生成内容免责声明

This app uses artificial intelligence to generate content. AI-generated content may be wrong, biased, incomplete, misleading, or otherwise unsuitable. The developer makes no representation or warranty regarding the accuracy, legality, suitability, completeness, reliability, or fitness of AI-generated content. You must independently review important information before relying on it.

AI-generated content is not medical, legal, financial, or other professional advice. Do not use AI-generated content as the sole basis for medical, legal, financial, employment, safety, or other high-risk decisions. For emergencies or high-risk situations, seek qualified professional help.

本应用利用人工智能技术生成内容。AI 生成的内容可能存在错误、偏见、不完整、误导性或其他不适用情形。开发者不对 AI 生成内容的准确性、合法性、适用性、完整性、可靠性或可用性作任何保证。对重要信息，您应当自行核验后再作依赖。

AI 生成内容不构成医疗、法律、金融或其他专业意见。请勿将 AI 生成内容作为医疗、法律、金融、就业、安全或其他高风险事项决策的唯一依据。如遇紧急情况或高风险场景，请及时寻求具备资质的专业人士帮助。

8. Age Requirement / 年龄限制

This app is published and intended only for users who are 18 years of age or older. If you are under 18, do not register for, access, or use the service.

本应用仅面向年满 18 周岁的用户发布和提供服务。如果您未满 18 周岁，请不要注册、访问或使用本服务。

9. Data Retention / 数据保留

We retain data only as needed to provide the service, maintain security, comply with applicable law, resolve disputes, and protect against abuse. If a user has not logged in or been active for 90 days, we delete, anonymize, or minimize retained data as part of implemented retention controls.

我们仅在提供服务、维护安全、遵守适用法律、解决争议和防止滥用所必需的范围内保留数据。如果用户连续 90 天未登录或未活跃，我们会按照已经实施的数据保留控制机制，对相关数据进行删除、匿名化或最小化保留处理。

10. Updates / 协议更新

We may update this agreement from time to time. The latest version will be available in the app. If you continue to use the app after an updated agreement becomes effective, you accept the revised version.

我们可能会不时更新本协议。最新版本会在应用内提供。如果更新后的协议生效后您继续使用本应用，即视为您接受修订后的版本。