Guide
This tool converts a stack of business card photos into a CSV or vCard file. Upload up to 20 cards at once; a vision AI reads each card and splits it into name, company, department, title, email, phone, website, and postal address. You review and correct the results against the original images, then download the file.
How to convert business cards to CSV
- Photograph each card straight-on in good light (one card per photo).
- Upload the photos — they are downscaled in your browser and stored privately in your workspace.
- Start the scan (2 credits per card, shown on the button before you run).
- Review rows marked "Check": click a row to compare against the original image and fix any field inline.
- Download the CSV (for spreadsheets or CRM import) or the vCard file (for your phone's contacts app).
Past runs are saved to your history, so you can re-open a batch and re-export it later.
What each card becomes
| Field group | Columns |
|---|---|
| Person | Full Name, Last Name, First Name, Name Reading (furigana/romaji) |
| Organization | Company, Department, Job Title |
| Contact | Email, Phone, Phone 2 / Fax, Website |
| Address | Postal Code, Prefecture, City, Address |
The AI is instructed never to guess: a field that is not printed on the card stays empty, and hard-to-read cards are flagged for review instead of being silently wrong.
Importing the CSV into a CRM
The CSV columns map directly onto common CRM contact properties:
| CSV column | HubSpot | Salesforce |
|---|---|---|
| First Name / Last Name | firstname / lastname | FirstName / LastName |
| Company | company | Account Name |
| Job Title | jobtitle | Title |
| Phone | phone | Phone |
| Prefecture / City / Address / Postal Code | state / city / address / zip | State / City / Street / PostalCode |
Example
A Japanese card that prints 「株式会社サンプル 営業部 部長 鈴木 良介 / [email protected] / 03-1234-5678 / 〒150-0002 東京都渋谷区…」 becomes one CSV row with Company = 株式会社サンプル, Department = 営業部, Job Title = 部長, Full Name = 鈴木 良介, plus the email, phone, postal code, prefecture, city, and street address in their own columns.
Limitations
- Handwritten cards and heavily stylized logo type reduce accuracy — always review flagged rows before importing anywhere.
- Vertical layouts and bilingual cards are supported, but unusual designs may need manual fixes.
- Just need the text off a single card? The free, in-browser business card OCR does that without signing in.
