Frequently Asked Questions
Got questions? We've got answers.
What kinds of documents work best in the JPG data extractor?
It shines on JPG photos and scans of paper documents with labeled fields — passports and other government-issued IDs, prescriptions and lab reports, health insurance claim forms, loan, mortgage, visa, and job applications, tax forms, and clinic intake or consent sheets. Anything with a printed label and a value next to it is fair game.
Does it work better with phone-camera photos or flatbed scans?
Both work. Flatbed scans give the most consistent results because the document sits flat, square, and is evenly lit. Phone photos work too — keep the document framed in the viewfinder, shoot from directly above, and avoid hard shadows. Modern phone cameras are usually high-resolution enough for clean extraction.
My JPG has glare, a shadow, or is slightly tilted — will it still extract correctly?
Yes, within reason. Moderate skew, soft shadow, and small glare patches are tolerated. Heavy glare across a critical field — like a hologram smear over a passport's date of birth — can hide the value. If a re-shoot is easy, diffuse the light (turn off direct flash, use natural daylight) and flatten the document before snapping again.
My phone took the photo sideways or upside-down (EXIF orientation) — does the extractor handle that?
Yes. JPGs from phones often carry EXIF orientation metadata so the preview displays correctly. The extractor reads the document at the right orientation regardless of how the raw pixels are stored, so sideways or rotated phone photos extract just like upright ones.
Are JPG and JPEG the same? Both supported?
Yes. JPG and JPEG are two extensions for the same JPEG image format — the only difference is the filename. Both are accepted as the primary input and processed identically.
What other formats besides JPG and JPEG can I upload?
PNG, TIFF, WebP, GIF, BMP, SVG, and PDF. If your scanner exports TIFF, your colleague sent a PDF, or your phone produced a HEIC-converted PNG, you can drop it in the same workflow without switching tools.
Is the JPG data extractor really free to use?
Yes — fully free for the core extraction features. No trial period, no credit card, no premium tier locked behind a paywall. Upload your JPG, extract, and export to JSON, CSV, Excel, or XML at no cost.
Do I need to sign up, log in, or install anything?
No. There's no account to create, no software to download, and no browser extension to install. Open the page, drop your JPG in, and go.
Should I use Auto mode or Manual mode for IDs, medical records, or application forms?
Use Auto when you're exploring a one-off document and want every label-value pair the AI can find. Use Manual when you're processing many copies of the same form — passport batches, intake sheets, claim forms — and need the exact same field names in every result. Manual locks the output schema so downstream systems get predictable columns every time.
Can I save a field template and reuse it across a stack of the same form?
Yes. In Manual mode, define your field names (and optional descriptions for accuracy), then download the template as JSON or CSV. Re-upload it for every future JPG of the same form — every batch of visa applications, every set of patient intake sheets — and the extraction produces a uniform schema.
Can I correct the extracted data before downloading it?
Yes. The result is shown in an editable table — fix a misread digit, correct a name, type in a value the AI missed, add a new field, or delete a row you don't need. Edits are applied before the file is generated, so the exported JSON, CSV, Excel, or XML reflects exactly what you confirmed.
What export formats can I download, and which fits which workflow?
JSON for APIs, scripts, and structured pipelines; CSV in horizontal layout (one row per document) for spreadsheet roll-ups, or vertical layout (key/value per row) for long-form reports; Excel in horizontal or vertical .xlsx for analysts who live in spreadsheets; XML for legacy systems and document-management tools. Pick whichever your destination expects.
What's the file size limit? My phone photos are several megabytes.
Max upload is 50 MB per file. That's plenty of room for typical phone-camera JPGs of single-page documents. If the photo is much larger, lowering camera resolution to around 8–12 MP, or saving at high JPEG quality (not maximum), usually shrinks the file enough without hurting extraction accuracy.
Does heavy JPEG compression hurt extraction accuracy?
It can. JPEG is a lossy format — at very low quality settings, small text and thin strokes blur into compression artifacts, and digits like 3/8 or 0/6 start to look alike. For best results, save your JPGs at high (not lowest) quality, especially for documents with small printed fields like IDs or lab reports.
Can it read handwritten entries on application or intake forms?
Printed and typed text gives the best accuracy. Clearly written, legible handwriting in form fields often works — block letters, dates, and signatures with printed names underneath are usually readable. Cursive, messy writing, or pencil on textured paper can drop accuracy; expect to spot-check those in the editable result table.
I scanned a multi-page form and got several JPGs — how do I extract from all of them?
The web tool processes one JPG at a time, so each page is a separate run — upload page one, extract, export; then page two, and so on. If you'd rather hand over a folder of JPGs and get a merged result, contact us about batch processing.
I'm uploading IDs and medical files — is my data private?
Yes. Files are processed for the extraction request and not stored, retained, or shared afterward. For organizations with compliance constraints — healthcare, financial services, government — contact us about local or on-premises deployment so JPGs never leave your infrastructure.
Can I run this in bulk, automate it, or get API access for a stack of IDs or claim forms?
The public web tool is one-JPG-at-a-time. For batch runs (hundreds of forms in one go), API access for embedding into KYC or claim-intake flows, CAPTCHA-free runs, multi-page or local pipelines, get in touch — we'll set up the right plan for the volume.