Pull Amount, Store Name, and Points from a PayPay Screenshot
PayPay generates a CSV export, but it drops the store name and the points earned — two details the app prominently shows on every payment confirmation screen. Screenshot extraction captures what the CSV chooses to leave out, in under 10 seconds per transaction.
5-10s per screenshot · Up to 99% accuracy on printed text
Fields a PayPay Screenshot Carries That CSV Loses
Five fields per confirmation — the CSV drops one entirely and generics another. Custom Column Extraction recovers both by reading content, not layout.
Bold ¥ figure on the yellow card — CSV stores it as a number but strips the currency context.
Shown with the merchant logo. The CSV also has it, but repeated visits become indistinguishable rows.
Absolute date or relative labels like "今日" — the AI returns what the captured screen offers.
A badge on the green completed screen — no equivalent column in the CSV export at all.
PayPay残高 / マネー / Card — immediate visual confirmation faster than parsing CSV columns.
A Clean Confirmation Screen with Two Quiet Gaps
PayPay's payment confirmation is well-designed — a clear ¥ figure, the store name with logo, a points badge, the payment method. The CSV export that launched in early 2025 covers the basics but makes two omissions that matter when you're tracking where your money went.
CSV Has No Points Column
Every PayPay confirmation screen shows the exact points earned — "○○ポイント獲得" — in a prominent badge on the green completed screen. The CSV export, despite covering 13 columns, has zero columns for points earned per transaction. The only trace of points appears as text buried inside the generic description field. On Reddit, users have asked "Is it possible to export transactions history in paypay, like we do for a bank account?" — years before any CSV feature existed.
Store Name Without Context
The CSV has a 取引先 (store name) column, but the 取引内容 (description) is always the generic "支払い" — payment. Three visits to the same convenience store in a week are three identical rows. The screenshot preserves the full merchant context.
Yellow UI vs. Standard OCR
PayPay's bright yellow confirmation card creates a low-contrast surface for traditional OCR. Semantic reading doesn't depend on contrast ratios; it reads content by meaning, not pixel color.
Extract Points Into Its Own Column
Define Points Earned as a column and the AI reads the badge from the green completed screen — something the CSV has no column for. Add a Computed Column to sum points across batches for monthly bonus tracking.
Store Name + Date + Amount in One Row
Define three columns — Store Name, Amount (¥), Date — and the AI returns them together per screenshot. The same three fields that appear on the yellow confirmation card become one spreadsheet row, with the points column that the CSV can't produce.
Semantic Reading Handles Yellow Backgrounds
The AI identifies text content by context — the ¥ amount sits in the center of the card, the store name sits at the top with a logo, the points badge sits on the green completion screen. It reads these relationships, not pixel colors, so PayPay's distinctive yellow interface doesn't reduce accuracy the way it would with traditional OCR.
From a PayPay Confirmation to a Spreadsheet Row
If you're logging PayPay transactions — for monthly household budgeting, tax filing preparation, or tracking points accumulation across multiple stores — here's what happens when you drop a screenshot into the tool.
Upload the Screenshot
Upload a PayPay screenshot — the yellow confirmation card before payment clears, the green "Payment Completed" screen with the points badge, the transaction history list showing relative timestamps, or the detail page with the absolute date. Any screen state works; the extraction reads content, not layout.
Name the Columns
Define your column headers: Amount (¥), Store Name, Date, Points Earned, Payment Method. Add an Inferred Column like Category (options: Groceries/Dining/Transport/Other) and the AI reads the store name to classify each transaction automatically — bypassing the CSV's generic "支払い" description entirely.
Export the Row
The output lands as one spreadsheet row — the ¥ amount, the store name (distinguishable from repeat visits), the date or relative timestamp, the points earned (which the CSV never captures), and the payment method. Export to Excel, CSV, or push directly into Google Sheets via the add-on. One screenshot, one row, no gaps.
PayPay Screenshot Extraction — When It Works and When to Look Closer
When It Works Best
Green completed screenshots. The store name, ¥ figure, and points badge all visible on a clean layout — the most reliable source.
Screenshots from the PayPay detail page. Tapping into a transaction reveals the full calendar date, store name, payment method, and transaction ID — the richest data density per screenshot.
Batch processing alongside other payment apps. The same column set extracts from PayPay, LINE Pay, WeChat Pay, and Venmo in one batch.
When to Be Cautious
Screenshots of the in-app transaction history list. Relative dates ("今日", "3日前") and truncated store names — the absolute date needs the individual detail page.
Compressed or forwarded screenshots. LINE and other apps compress shared images. The AI still outperforms traditional OCR on yellow backgrounds, but forwarded copies are worth spot-checking.
Pending or in-progress transactions. The yellow "処理中" card shows amount and store but the points badge hasn't been awarded. Wait for the green completed screen.
Frequently Asked Questions
Can I extract the points earned from a PayPay screenshot when the CSV doesn't include them?
Yes. The green "Payment Completed" screen displays the exact number of PayPay points earned as a badge — for example "5ポイント獲得". Define a Points Earned column and the AI reads this value directly from the badge, even though the CSV export has no equivalent column. This is the single most valuable field that screenshot extraction recovers that the official export omits.
Does the extraction work on PayPay's bright yellow confirmation screen?
Yes. Traditional OCR engines rely on high-contrast edges between text and background — something PayPay's signature yellow (#FFD600) interface does not provide for dark text. The AI extracts by reading content semantically, understanding that the large centered figure is the Amount (¥) and the label at the top identifies the Store Name, regardless of the background color. The yellow confirmation card and the green completed screen are both handled without adjustment.
Can I batch PayPay screenshots with LINE Pay or WeChat Pay screenshots in the same export?
Yes. Define one set of column names — Amount, Store Name, Date, Points Earned, Payment Method — and the AI processes PayPay, LINE Pay, WeChat Pay, Venmo, and PayPal screenshots in a single batch using the same definitions. Each screenshot becomes one row in the output spreadsheet, regardless of which app generated it. See the payment screenshots overview for the full cross-app capability.
Does the tool support PayPay's relative timestamps like "今日" (Today) or "3日前"?
The AI reads whatever date information is visible on the screen. If the screenshot is from PayPay's transaction history list showing "昨日" or "3日前," that relative string is what goes into the Date column — it's a useful rough reference but not a permanent record. For the absolute calendar date — month, day, year, and precise timestamp — always screenshot the individual transaction detail page by tapping into the payment from the history list. The AI handles both inputs transparently.
Can I distinguish between PayPay Balance and PayPay Money in the extracted Payment Method field?
Yes. The confirmation screen shows the exact funding source — PayPay残高 (Balance), PayPayマネー (Money), or PayPayカード (Card) — and the AI reads this label into the Payment Method column. This is particularly useful for users who maintain both a PayPay Balance and a PayPay Money wallet, as the CSV export does include a payment method field but the screenshot provides immediate visual confirmation without needing to cross-reference column data.