If you proactively use the official AI Pro service for bill recognition, receipt parsing, or structured posting, we process information related to that request, including:
如果您主动使用“AI Pro 服务(官方)”进行账单识别、票据解析或结构化入账,我们会处理与该次请求相关的信息,包括:
- Bill text, OCR text, or speech-to-text content that you enter or upload您输入或上传的账单文本、OCR 文本、语音转写文本
- Bill images or compressed image data that you submit for recognition您主动提交用于识别的账单图片或压缩后的图片数据
- Prompt content, template information, and model routing parameters needed to complete the recognition为完成识别所需的提示词内容、模板信息、模型路由参数
- Client identifier
client_id客户端标识 client_id
- Necessary fields related to quota, entitlements, risk control, and request source与额度、权益、风控、请求来源相关的必要字段
We process this information to provide AI bill recognition and structured output, determine available models and entitlement status, perform billing, rate limiting, abuse prevention and service security, and conduct necessary troubleshooting and service stability maintenance.
处理目的包括:提供 AI 账单识别与结构化返回结果,判断可用模型、额度和权益状态,进行计费、限流、反滥用和服务安全保障,以及进行必要的故障排查与服务稳定性维护。
In this mode, the typical data-processing chain is your device, our server-side interface, and third-party large-model or AI service providers integrated by us.
在该模式下,相关数据处理链路通常为:您的设备、我们的服务端接口,以及我们接入的第三方大模型或 AI 服务商。
Based on the current project configuration, such third-party model providers may include but are not limited to DeepSeek, Alibaba Tongyi Qianwen / DashScope, Tencent Hunyuan, Zhipu AI, and other compatible providers added in the future.
根据当前项目配置,第三方模型服务商可能包括但不限于:DeepSeek、阿里云通义千问 / DashScope、腾讯混元、智谱 AI,以及其他后续接入的兼容模型服务商。
To avoid unnecessary disclosure of sensitive information, please do not upload highly sensitive information unrelated to bookkeeping, such as full ID numbers, full bank card numbers, medical information, precise home addresses, or other content beyond what is necessary for bookkeeping.
为避免敏感信息不必要扩散,请勿上传与记账无关的高敏感信息,例如完整身份证号码、银行卡完整卡号、医疗信息、精确住址或其他超出记账目的所必需的信息。