Scan any business card — even if you don't understand the language
You just met someone. Their card is in Hindi, Arabic, or Japanese. You don't read it — but you still need to follow up.
Why most business card scanners fail outside English
- ✕Latin-only OCR mangles non-Latin scripts
- ✕Mixed scripts (English + Hindi on the same card) confuse most apps
- ✕Right-to-left languages (Arabic, Hebrew, Persian, Urdu) get reversed or skipped
- ✕Non-standard layouts get misread as a single block
- ✕Result: broken names, missing fields, useless contacts
How NexaLink reads any business card
1
Scan the card
Open camera, frame card, tap.
2
Language auto-detected
No picker. NexaLink identifies the language(s) present in <500ms.
3
Saved to contacts
Native script preserved. Mixed-language cards keep both scripts intact.
Where it actually matters
International conferences
You meet 20 people from 10 countries — every card scanned, no confusion later.
Read more →Works across 38 languages
- Indian languages (12): Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Urdu, Odia, Assamese
- Middle East (5): Arabic, Hebrew, Persian, Swahili, Amharic
- Europe (14): Italian, Dutch, Russian, Polish, Turkish, Ukrainian, Greek, Czech, Hungarian, Romanian, Swedish, Norwegian, Danish, Finnish
- ASEAN (7): Indonesian, Vietnamese, Thai, Malay, Tagalog, Burmese, Khmer
- East Asian (existing): Japanese, Chinese, Korean
- Western (existing): Spanish, French, German, Portuguese
See it on a real card
“Reads mixed Hindi+English cards without breaking formatting”
asset pending
Sample Hindi+English business card
800x500 photo or render of the test card
Card with English name + Hindi address, mixed-script layout
Before
Generic OCR: name fragmented, Hindi block dropped
After
NexaLink: name + Hindi address captured cleanly, both scripts preserved
Directional — based on internal QA across 50 mixed-language test cards. Not third-party certified.
Auto-detect, then verify
- Detection happens on-device for privacy
- User can override the detected language with one tap
- Native script preserved in the contact record
Frequently asked
Can it scan mixed-language cards?▾
Yes. A card with English on one side and Hindi (or Arabic, or Chinese) on the other is fully captured — both scripts go into the saved contact record without you having to choose between them.
Does it translate or preserve the original script?▾
It preserves the original script. The address in Hindi stays in Hindi, the name in Arabic stays in Arabic. We don't transliterate without your consent because transliteration loses information.
Does it work offline?▾
Language detection runs on-device, so basic capture works offline. The full save-to-contact flow needs a connection for the first scan; cached scans sync once you're back online.
Can I export to a CRM?▾
Yes. NexaLink contacts export to Salesforce, HubSpot, and a generic CSV. Mixed-language fields export as-is — your CRM may or may not handle non-Latin scripts well, depending on its locale.
How accurate is the Hindi / Arabic / Chinese OCR?▾
Directionally: name + phone + email are captured correctly on >90% of well-printed cards. Address fields and handwritten elements are lower. We're conservative on accuracy claims because card quality varies enormously.
What happens with handwritten cards?▾
OCR handles printed text well; handwriting is unreliable across all apps including ours. If a card is handwritten, you'll see the extracted text in a confirmation step where you can correct before saving.
Privacy — where do my scans live?▾
On-device first. The scan image and extracted text sync to your encrypted NexaLink cloud only after you save the contact. You can delete a scan; the original image is purged within 7 days of deletion.
iOS only?▾
iOS is fully supported with all 38 languages. Android supports the same 38 languages on devices running Android 10+; older devices fall back to a smaller set.
Stop typing business cards manually
Scan → Save → Follow up faster