Bancha is built around the idea that your data stays with you. We do not track you, we do not collect your personal data on our servers, and there are no ads or third-party analytics.
What the app stores, and where
Your tracking content is stored on your iPhone/iPad:
Watch progress for anime, per-rewatch history, per-episode dates and notes, ratings
Your profile (nickname, bio, avatar photo)
Unlocked achievements
App settings (display language, etc.)
Starting with Bancha 1.3, when iCloud Sync is on (Settings › Data, on by default), this data is also stored in your own iCloud account (Apple's CloudKit private database and iCloud key-value storage) so that iPhones and iPads signed in to the same Apple Account stay in sync. Data in iCloud is held by Apple and can be read only by your Apple Account; we cannot see it. When you turn iCloud Sync off, this device stops exchanging data with iCloud.
This data is not sent to our servers. You can delete any entry in the app. Settings › Data › Clear All Data can clear this device only, or clear both this device and iCloud. Deleting the app does not delete the data in iCloud; you can also delete Bancha's iCloud data in iPhone Settings › Apple Account › iCloud › Manage Account Storage. The app also offers Export (JSON/CSV) so you can back up your own data.
Bancha 1.2 and earlier has no iCloud sync. Data is stored only on the device, and deleting the app removes all of it.
Linking AniList (Bancha 1.2 and earlier, optional)
Bancha 1.3 and later has no sign-in of any kind and no AniList link.
In Bancha 1.2 and earlier, if you choose to link your own AniList account, the app syncs your tracking data with your AniList account via AniList's official authorization (OAuth). This happens only after you authorize it; data flows between your device and AniList and is governed by AniList's Privacy Policy. Your AniList access token is stored in your device's Keychain. You can unlink at any time in-app.
What our server receives
Bancha has a helper server whose only purposes are to provide Traditional Chinese titles and synopses for works, basic service status, and crash fixing. It only receives:
A work's ID (to look up its Traditional Chinese title, synopsis, and airing times)
The ID of a work that has no Traditional Chinese title yet (so we can add the translation later)
Requests to download the work catalog (Bancha 1.3 and later; these requests carry no personal data. Search looks up that catalog on your device, and your search keywords are not sent)
Bancha 1.2 and earlier: the search keywords you type when you use Search (used to look up matching Traditional Chinese titles)
A one-time service-status check when the app launches (used to show maintenance notices or critical update reminders; this request carries no data)
When you choose to report a data problem, what you filled in (work ID, problem type, your suggested title)
When you tap Report This Title after a search found nothing, the keywords you had typed
When the app crashes or stops responding, a diagnostic summary (app version, build number, iOS version, device model identifier, event type, a hash of the call-stack summary, and the date)
The server does not receive or store any information that identifies you (no account, no name, no email, no device identifier, none of your watch history).
Other services the app contacts directly
Bancha 1.3 and later: the app contacts Bangumi (bgm.tv) directly for episode lists, cover art, and genre rankings. Those requests carry only a work ID or a genre tag.
Bancha 1.2 and earlier: the app contacts AniList directly for work details and cover art.
Used only for that one request
Work-ID lookups, catalog downloads, service-status checks, and the search keywords sent by version 1.2 and earlier are used only for that single request. The server keeps none of that content, and our code does not log your keywords. As with any ordinary network request, the server sees your IP address at the moment of connection; we do not use it to identify you or build any profile.
Cloudflare, which hosts the server, retains connection logs for a period of time. Those logs include the request URL and are used for abuse prevention and debugging. Search keywords (1.2) and search keywords you choose to report are sent in the request body, so they do not appear in the URL.
What is kept, and for how long
Data problems you report (work ID, problem type, your suggested title): kept for 180 days, used only to correct Traditional Chinese titles and synopses, then deleted automatically.
Search keywords you report: kept for 180 days, used only to add missing titles, aliases, and works, then deleted automatically. The content is that text and a timestamp.
IDs of works with no Traditional Chinese title, and how often they were requested: kept until we add that title, used only to prioritize translation work. The content is an ID and a count.
Crash and hang diagnostic summaries: kept for 90 days, used only to fix crashes. The content is the items listed above; the call stack is hashed on your device before it is sent, so the server never sees the stack itself and receives no text you typed. You can turn diagnostics off in the app's settings ("Share Diagnostics").
Abuse-prevention counters: counted against a hashed source address, kept for at most 1 hour, used only to stop flooding of the report endpoints. The hash changes every minute and cannot be correlated across time.
What we do not do
No cross-app/cross-site tracking (the app's Privacy Manifest declares NSPrivacyTracking = false)
No device identifiers (IDFA), no advertising
No third-party analytics, marketing, or social SDKs
We never sell or share your data
Children
This app is general-audience, not directed at children under 13, and does not knowingly collect children's personal data.
Changes to this policy
If this policy materially changes, we will update the "Effective date" on this page.