JPG2Excel converts photographs of tables into working spreadsheets: point it at a JPG of a price list, a printed report, a bank statement or a menu, and its OCR detects the rows and columns and returns an Excel file instead of a wall of undifferentiated text.
The row-and-column detection is the entire product. Plain OCR has been free for a decade; what it gives you is a paragraph of numbers with the table structure destroyed. Rebuilding structure is the part that eats an afternoon, and the part this tool is selling back.
What it costs
Free to start, with paid volume in the $7–48 range. For occasional table-rescue that free tier may simply be enough; the paid tiers are aimed at people digitising stacks, not pages.
What we can’t tell you
Accuracy on the tables that actually get photographed — skewed phone shots, merged cells, columns that wrap. And one instruction we will state as policy rather than suggestion: verify extracted numbers before they touch anything financial. OCR errors in digits are silent, plausible and expensive; a transposed figure looks exactly like a real one. Spot-check totals against the source every time.
The details
| What it is | OCR for tables: photo in, structured Excel out |
| Price | Free to start · paid volume ~$7–48 |
| Best for | Digitising printed tables without retyping the structure |
| Think twice if | The numbers feed accounts unchecked — always verify digits against the source |
