Every paper in E-Porikkha starts with student text. There are two ways to get that text in: paste it directly, or let the OCR engine extract it from a scan. Both paths lead to the same grading studio.
Option 1 — Paste text (recommended for typed answers)
If students submit answers digitally — via a Google Form, a Word document, or any typed format — just copy the text and paste it into the upload modal. No OCR step, no waiting, no extra cost.
This is also the best path if you're processing a backlog of already-transcribed papers.
Option 2 — Upload a scan
Upload a JPEG, PNG, or PDF of the handwritten script. E-Porikkha runs OCR in the background (via Celery, asynchronously) and shows you a spinner. Typical extraction time is 10–30 seconds per page.
Once extraction finishes you can edit the text before grading — useful for fixing any [illegible] markers or obvious mis-reads.
Credit usage
Credits are charged based on the real AI cost of what actually ran on your paper — not a flat fee per paper. Pasted text skips OCR entirely, so it only ever costs whatever grading itself uses. A scanned upload additionally costs a small OCR fee (billed per page), so a 3-page script and a 25-page script are no longer priced the same.
Your very first regrade on a paper is free; every regrade after that is billed the same way, based on what it actually costs to re-run.
Which should I use?
Use paste whenever possible — it's faster and slightly cheaper, since it skips the OCR step entirely. Use scan when you have physical scripts that haven't been transcribed.
Start grading with 2 free credits — no credit card required.