ReceiptClaim Privacy Policy

Effective date: August 22, 2026

ReceiptClaim is a receipt splitting app published by TinyNeed. It creates no account, shows no ads, tracks no location, touches no payment account, and sells no personal data. Everything below describes the only ways receipt data is handled.

Cloud recognition is optional. Because it sends and retains data outside your device, the cloud and on-device paths are described separately, and the app tells you which one it is about to use. Anyone who accepted the earlier non-retention disclosure must accept the new retention disclosure before another cloud scan is sent.

On-device recognition

When you choose or photograph a receipt, the app can recognize the text on your device using Google ML Kit. In ReceiptClaim 1.1 and earlier this is the only recognition path, and no receipt image or recognized text is sent anywhere.

ML Kit may send Google limited device and app information, a per-installation identifier, performance metrics, API configuration, feature input and output sizes, feature versions, event types, and error codes. On Android and iOS, Google describes these metrics as used for diagnostics and usage analytics, states that input images, input text, and recognition results stay on the device, and states that the per-installation identifiers are not intended to uniquely identify a user or physical device. Google states that the Android data is encrypted in transit and is not shared with third parties. ReceiptClaim does not use these metrics for advertising or tracking. See Google's ML Kit Android data disclosure, ML Kit iOS data disclosure, and ML Kit terms.

Cloud recognition

Cloud recognition reads harder receipts more accurately than a phone can. The app explains the retention below before it would use the cloud. If you decline, that choice is remembered and no receipt image is sent; on-device recognition remains fully available, free, unlimited, and usable offline.

The app offers Automatic, Mainland China, and Global routing. Automatic may choose the mainland route when the device supplies a mainland China region hint. You can override that choice in the app. A failed Global request is never sent to the Mainland China service as a fallback.

Global route. The receipt image is sent over an encrypted connection to the ReceiptClaim Cloudflare service. The app-owned copy is stored in Cloudflare R2. Recognition requests are routed through OpenRouter to the configured AI model providers, currently Google and OpenAI. OpenRouter says prompt and response logging is off by default, but it stores request metadata and may process files for routing, abuse prevention, security, billing, or legal compliance. Model providers have their own data practices.

Mainland China route. The receipt image is sent over an encrypted connection to rc.gojito.top. The app-owned copy is stored on the ReceiptClaim server in mainland China, and recognition is performed by Alibaba Cloud Model Studio (Bailian/Qwen) in a dedicated mainland China workspace. Alibaba Cloud states that Model Studio does not use customer data for model training and that the selected region determines its access and storage location. See Alibaba Cloud's Model Studio privacy notice and region documentation.

For either cloud route, ReceiptClaim stores the receipt photo, the raw recognition response and parsed result, the app-generated Device ID, a scan reference, and route, model, timing, outcome, and request metadata. We use this archive to investigate recognition errors, improve accuracy, and respond to support and deletion requests.

We do not send participant names, item claims, adjustments, or settlement data to cloud recognition. Those remain on your device.

What we store

To count cloud scans against your monthly allowance, the app generates a random Device ID the first time cloud recognition is used and stores it on your device. It is not derived from your hardware and carries no meaning outside ReceiptClaim. Our service stores that identifier with scan counts, purchase entitlements or credits, and each retained cloud-scan archive. The recognition model does not need the Device ID and does not receive it as part of the recognition prompt.

Local receipt drafts, item claims, participant names, adjustments, and history stay on your device.

Website analytics

This website is hosted and protected by Cloudflare and uses Cloudflare Web Analytics to measure aggregate site traffic and performance. It does not accept or upload receipt photos, recognized receipt text, participant names, or settlement data, so that content is not part of website analytics. Cloudflare may process website request and analytics data under its privacy policy.

Camera and photo access

Camera and photo-library permissions are used only when you ask the app to capture or choose a receipt. The selected image is read by on-device recognition or, after the cloud-retention disclosure is accepted, sent and stored for the cloud reading described above. No other image is accessed, and the limited ML Kit metrics described above never contain receipt content.

Storage and deletion

The current draft remains on your device until you start another receipt, clear the app's storage, or uninstall the app. Starting another receipt clears the saved draft inside ReceiptClaim.

Cloud-scan archives currently have no automatic expiry. They are retained until removed for service operations or in response to a valid privacy request. To request deletion, email [email protected] and include the Device ID shown in Support info or the scan reference shown with the affected scan. Do not attach the receipt itself unless you are comfortable sharing it by email.

Sharing

If you use the Share action, the settlement text is passed to the destination you choose through the operating system's share sheet. The chosen third-party app then handles that text under its own privacy policy.

Children

ReceiptClaim is a general utility and is not directed to children under 13. We do not knowingly collect children's personal information.

Changes

If the app's data practices change, we will update this page and its effective date before the new practice applies.

Contact

Questions or privacy requests: [email protected]