隐私政策
翻丫（TransUtter，以下称“本 App”）在简体中文、英语、法语、西班牙语之间互译播客、网页文章和文本，并用语音朗读出来。这份政策说明我们收集什么、为什么收集、会交给谁处理，以及你怎么删除它。
三句话版本：
- 我们不使用任何第三方分析、广告或跟踪 SDK，不收集你的通讯录、位置或设备标识符用于广告。
- 选择“本机”处理时，你的文本和音频不离开这台设备。
- 选择“云端”处理时，你提交的内容会上传到我们的服务器，并交由受我们委托的第三方云服务处理。
一、账号信息
本 App 只提供“通过 Apple 登录”（Sign in with Apple）一种登录方式。登录时 Apple 会给我们一个针对本 App 的匿名用户标识符。
- 我们保存：该匿名标识符、我们生成的内部账号 ID、账号创建时间、账号状态。
- 我们不保存你的邮箱地址。即使你在登录时选择共享真实邮箱，我们的服务端也只用它完成本次登录校验，不写入数据库。
- 我们拿不到你的 Apple ID 密码，也拿不到你的姓名（除非你主动写在反馈邮件里）。
二、你提交的内容
本 App 有两种处理方式，隐私影响完全不同：
本机处理
文本翻译与本地音频转写在你的设备上完成，使用 Apple 的翻译框架与设备端语音识别模型。这些内容不会上传到我们的服务器，也不会发给任何第三方。首次使用时设备会下载所需的语音识别模型文件，这个过程不包含你的任何内容。
云端处理
播客转译以及你主动选择云端时，以下内容会上传到我们的服务器：你粘贴的文本、你分享的网页文章正文、你上传的音频文件、你提交的播客链接及其对应的音频。
为完成语音识别、机器翻译和语音合成，上述内容会交由受我们委托的第三方云服务处理。这些服务商仅按我们的指示处理数据，用途限于提供本 App 的转换功能，不得用于其他目的。我们会随技术条件变化调整所使用的服务商。
这些服务商可能依据其所在地的法规对内容进行合规审核，因此个别内容可能无法完成处理。发生这种情况时，App 会向你说明并且不扣除点数。
处理过程中产生的临时文件在任务结束后即被删除。
三、我们不保存你提交的内容
你上传的音频文件、你粘贴的文本、你分享的网页文章正文，以及它们的处理结果，我们都不长期保存。这些内容只在转换过程中存在于我们的服务器上，转换完成并由你的设备取回后即被删除。
- 用于处理的原始上传文件在任务结束时即从服务器删除。
- 处理结果（转写文稿、译文、合成音频）在你的设备取回后删除，不会保留在我们的服务器上。
- 我们不会用你提交的内容训练模型。
如果你对某条内容有额外的删除需求，请通过下方邮箱联系我们。
四、购买信息
订阅通过 App 内购买（Apple 的 App Store）完成。支付全程由 Apple 处理，我们不接触也不保存你的银行卡号、支付信息或账单地址。
为了给你的账号发放点数、处理续订与退款对账，我们保存 Apple 提供的交易记录：交易 ID、商品 ID、购买时间、购买环境（正式/沙盒）、是否已退款，以及 Apple 返回的原始交易凭证。
五、通知
本 App 只使用本地通知（转换完成提醒、订阅播客的新集提醒）。新集检查在你的设备上进行，我们不收集推送令牌，也不向你的设备推送服务器消息。
六、我们不做的事
- 不集成第三方分析、崩溃统计或广告 SDK。
- 不收集广告标识符（IDFA），不做跨 App 跟踪。
- 不访问你的通讯录、相册、位置或健康数据。
- 不出售你的个人信息，也不把你的内容用于训练我们自己的模型。
七、数据保留与删除
- 删除账号：App 内“设置”里可直接删除账号。执行后我们会删除你的登录凭据并将账号标记为已删除，你将无法再用它登录。
- 额度账本会保留：出于财务审计和退款对账的需要（这也是 Apple 的要求），与交易和额度变动相关的最小记录会被保留，其中不含你的姓名或邮箱。
- 未使用的点数在账号删除后不予退还，也无法转移。删除账号不会自动取消订阅，请另行在 iOS 的订阅设置中取消。
- 本机数据：删除 App 即删除设备上的全部历史记录、缓存结果与已下载模型。
八、儿童
本 App 不面向 13 周岁以下儿童，也不会有意收集儿童的个人信息。
九、政策变更
政策更新时我们会修改本页顶部的生效日期。涉及重大变更时，我们会在 App 内提示。
十、联系我们
隐私相关问题、数据删除请求，请邮件联系：support@petlec.com。
Privacy Policy
TransUtter translates podcasts, web articles and text between Simplified Chinese, English, French and Spanish, and reads the result aloud. This policy describes what we collect, why, who processes it, and how you delete it.
The short version:
- We use no third-party analytics, advertising or tracking SDKs, and collect no contacts, location or advertising identifiers.
- In on-device mode, your text and audio never leave your device.
- In cloud mode, what you submit is uploaded to our server and processed by third-party cloud services acting on our instructions.
1. Account information
Sign in with Apple is the only sign-in method. Apple gives us an app-specific anonymous identifier.
- We store: that anonymous identifier, an internal account ID we generate, the creation time and the account status.
- We do not store your email address. Even if you choose to share your real address at sign-in, our server uses it only to complete that sign-in and never writes it to the database.
- We never receive your Apple ID password or your name.
2. Content you submit
On-device processing
Text translation and local audio transcription run on your device using Apple's translation framework and an on-device speech-recognition model. This content is not uploaded to us and not shared with any third party. On first use the device downloads the speech-recognition model files; that download contains none of your content.
Cloud processing
For podcast conversion, and whenever you choose cloud mode, the following is uploaded to our server: text you paste, article text you share, audio files you upload, and podcast links plus the corresponding audio.
To perform speech recognition, machine translation and speech synthesis, that content is processed by third-party cloud services engaged by us. They act only on our instructions, solely to provide the app's conversion features, and for no other purpose. The providers we use may change as the technology does.
Those providers may screen content for compliance with the regulations that apply to them, so some submissions cannot be completed. When that happens the app tells you, and no minutes are charged.
Temporary files created during processing are deleted once the job finishes.
3. We do not keep what you submit
Audio files you upload, text you paste, article text you share, and the results produced from them are not kept. They exist on our servers only while the conversion runs, and are deleted once it finishes and your device has retrieved the result.
- The uploaded source file is deleted from the server as soon as the job finishes.
- Results (transcript, translation, synthesised audio) are deleted once your device has retrieved them. They are not retained on our servers.
- We do not use what you submit to train models.
If you have a further deletion request for specific content, contact us at the address below.
4. Purchase information
Minute packs are bought through Apple's App Store. Apple handles payment end to end; we never see or store card numbers, payment details or billing addresses.
To credit minutes and reconcile refunds we store the transaction data Apple gives us: transaction ID, product ID, purchase time, environment (production/sandbox), refund status and Apple's signed transaction receipt.
5. Notifications
The app uses local notifications only (conversion finished, new episode of a followed podcast). New-episode checks run on your device; we collect no push tokens and send no server-initiated messages.
6. What we don't do
- No third-party analytics, crash-reporting or advertising SDKs.
- No advertising identifier (IDFA), no cross-app tracking.
- No access to contacts, photos, location or health data.
- We do not sell personal information, and we do not use your content to train our own models.
7. Retention and deletion
- Account deletion is available in the app's Settings. We delete your sign-in credential and mark the account deleted; it can no longer be used to sign in.
- The quota ledger is retained: for financial audit and refund reconciliation (also an Apple requirement) we keep the minimal records of transactions and quota changes. They contain no name or email.
- Unused minutes are not refundable or transferable after account deletion.
- On-device data: deleting the app removes all local history, cached results and downloaded models.
8. Children
TransUtter is not directed at children under 13, and we do not knowingly collect their personal information.
9. Changes
We will update the effective date at the top of this page when this policy changes, and notify you in the app for material changes.
10. Contact
Privacy questions and deletion requests: support@petlec.com.