UJI Pen Characters (Version 2)
Prat, F., Castro, M., Llorens, D., Marzal, A., & Vilar, J. (2008). UJI Pen Characters (Version 2) [Dataset]. UCI Machine Learning Repository.
Pricing
Your handwriting is free to turn into a font, and free to keep. You only pay when something goes in the mail.
Sending is in test mode. Letters and postcards are rendered as PDF proofs and nothing is printed or mailed yet. The prices below are what a send will cost once mailing opens.
Prices
Fonts
Free
- Make fonts in quick or full mode
- Download your fonts as TrueType, as often as you like
- Download any library font
- Use them anywhere, commercially too
Handwritten letter
$1.00
or 100 credits
- Your text, written in your font
- Printed, enveloped and mailed for you, once mailing opens
- A proof and tracking of its status
Handwritten postcard
$1.65
or 165 credits
- A short note in your hand on one side
- The address on the other
- A proof and tracking of its status
Prices in US dollars, including US postage; US addresses only for now. Paid with credits; card payments are not switched on yet.
Credits
Credits are a thank-you for making things here. You earn them without paying and spend them on mail. One credit is worth $0.01 toward a send.
Earn
| Make your own fontfor your first complete font (all 26 letters) | +10 credits |
| Download a library fontonce per library font, up to 5 a day | +2 credits |
| Send a letter or postcardback on every send that is mailed | +5 credits |
Spend
| A handwritten letter | 100 credits |
| A handwritten postcard | 165 credits |
Each font and each library download earns once, so re-downloading doesn’t add up. Your balance and every entry behind it are on your account page. Credit amounts may change before launch.
Attribution
The library fonts, the sample hand on our home page, and the model that suggests letters in quick mode all come from handwriting that researchers recorded and shared openly. We credit them here and inside every font and letter that uses their data.
ISGL Online and Offline Character Recognition Dataset (Version 1)
Adeyanju, I., Omidiora, E., Fenwa, O., Babalola, O., & Durodola, O. (2019). ISGL Online and Offline Character Recognition Dataset (Version 1) [Dataset]. Mendeley Data.
What we changed. We selected samples per writer, normalised and resampled the pen strokes, placed them on a baseline, smoothed them and converted them to font outlines; the suggestion model was trained on these strokes only. No endorsement by the dataset authors is implied.
These credits cover the dataset material. They don’t apply to your own handwriting: letters you draw are yours.