Convert a Picture to Excel on iPad (no retyping)

The table you need is an image on your iPad. A screenshot of a dashboard that has no export button. A price list a supplier attached to an email as a JPEG. A photo of a printed schedule that synced over from your iPhone. A rate table someone dropped into a group chat as a picture of their spreadsheet. You can read every cell — and none of it is usable. Long-press gives you Live Text, which selects the words but throws away the columns. Paste into Numbers and you get one column of fragments. Most “convert to Excel” tools won’t take the file at all, because a picture isn’t a PDF.

This guide is how to convert a picture to Excel on iPad in 2026: take the image exactly as it exists — screenshot, JPEG, HEIC, PNG — share or drag it into an AI extractor that reads the image the way you do, and get back a real .xlsx with the columns separated and the numbers typed as numbers. On an M-series or A17 Pro iPad running iPadOS 26, the image never leaves the device.

Why the iPad’s built-in tools can’t finish the job

iPadOS reads images extremely well and hands you tables not at all:

Live Text extracts words, not structure. Long-press any image of a table and the text selects — as lines of words in recognizer order. The geometry that made it a table (this price belongs to that part number) is discarded the moment you copy. What lands in Numbers is dates, then descriptions, then amounts, interleaved.
There is no “export as table” anywhere in iPadOS. Photos can lift a subject out of an image, translate a sign, identify a plant. Through iPadOS 26 there is no path — in Photos, Files, or the share sheet — from a photographed table to a spreadsheet.
Screenshots have the same problem as photos of paper. A screenshot of an Excel file in a chat, a web dashboard, or a locked PDF preview is pixels. The source being digital doesn’t help; the structure didn’t survive the screenshot.
Web converters mostly want PDFs. The ones that accept images run classic OCR: recover characters and positions, then guess column boundaries from pixel gaps. On a real photo — perspective, a shadow, uneven light — the guesses collapse into fragment soup. And you uploaded the image to get that result.

The gap is specific: the iPad can read the characters, but nothing built in understands the table.

What actually works: AI that reads the image, on the iPad

ignitai takes images directly — HEIC, JPEG, PNG, screenshots — no PDF wrapper required. It reads the image the way a person does: recognizes the table as a table, headers and rows and the relationship between a quantity and its line total, and writes that understanding straight into a spreadsheet. Character recognition and structure understanding happen in one step, so there’s no separate OCR pass whose column-guessing errors you inherit.

The full iPad flow:

Get the image into ignitai. Two equally good paths, and the second is the one the iPad was built for:
- Share sheet: from Photos, Mail, Messages, or Files — tap Share → ignitai. Select several images at once if the table spans pages.
- Drag and drop: open ignitai in Split View or Stage Manager next to Mail, Safari, or Photos, and drag the image straight across. For images embedded in an email or a web page, dragging skips the save-to-Photos detour entirely.
Say what you want, in plain English. For an image that is mostly table — a price list, a schedule, a rate card — skip the prompt and let ignitai take everything. For a busier image, one sentence picks the data out of the noise:

“Extract the rate table: plan name, monthly price, annual price. One row per plan. Ignore the navigation bar and the cookie banner.”

Prompts matter more on images than on clean PDFs, because images contain the world — browser chrome around a screenshot, a thumb at the edge of a photo, Apple Pencil circles someone drew on the page. Naming what you want tells the model what to ignore. Save the prompt as a preset for next time.
Choose Excel (.xlsx) as the output. Column types are set during extraction: dates become dates, prices become numbers with currency formatting. The sheet opens in Numbers or Excel for iPad ready to sort and sum — not as text pretending to be numbers.
Tap Extract. On an M-series or A17 Pro iPad running iPadOS 26, the extraction runs entirely on-device via Apple’s Foundation Models framework — the image is read on the iPad and the .xlsx is written on the iPad. A single image takes seconds. On older hardware or iPadOS versions, ignitai falls back to a hosted pipeline with documented zero retention.
Verify in Split View, then save. This is where the iPad beats the iPhone at this job: put the original image on one side of the screen and the extracted sheet on the other, and spot-check the first and last rows with both in view. Save to Files, iCloud Drive, or share it onward.

Total user time: under a minute from image to .xlsx.

Screenshots: the iPad’s most common table-in-a-picture

On iPhone, the typical input is a camera photo. On iPad, it’s a screenshot — and that changes the workflow slightly in your favor:

Screenshots are the easy case. Machine-rendered type, perfectly flat, evenly lit. They extract essentially without error. The dashboard with no export button, the pricing page you want in a comparison sheet, the spreadsheet screenshot someone posted instead of the file — all one share away from being real data.
Long tables: overlap and stack. If the table is longer than one screen, take overlapping screenshots as you scroll, select them all in Photos, and share them as one job. ignitai writes one sheet; delete the duplicate boundary rows in the preview. (For a web page, also check whether Safari’s full-page screenshot offers itself — it saves as PDF, which works just as well: same app, same prompt.)
Crop is optional, prompt is easier. You can crop browser chrome out of a screenshot before sharing it, but a prompt clause like “ignore everything outside the table” does the same job with fewer taps and lives in your saved preset.

Photos of paper: usually taken elsewhere, converted here

Most paper gets photographed with the iPhone in your pocket, not the iPad on your desk — and then the shot is already in the iPad’s Photos library via iCloud before you sit down. That’s the natural division of labor: capture on the phone, convert and verify on the bigger screen. When you are shooting with the iPad’s own camera, the same rules apply as anywhere: square-on, no shadow across the page, table filling the frame.

Two boundary cases worth routing to their own workflows:

Multi-page paper documents. If the paper matters and runs past a couple of pages, capture it with the Files or Notes document scanner instead of the camera — it deskews and normalizes lighting, and produces a multi-page PDF. From there, the scanned-PDF-to-Excel workflow on iPad is the more durable habit.
Receipts. Crumpled thermal paper is a solved but specifically-handled problem — the receipt-to-spreadsheet workflow on iPad covers capture to expense sheet end to end.

The picture path in this guide is for the images that already exist: the screenshot, the texted JPEG, the photo someone else took.

Batches: many images, one sheet

The realistic version of this job is rarely one image. It’s nine screenshots of a competitor’s pricing pages, or a month of delivery manifests texted as photos, or fourteen receipt shots from a trip:

In Photos, select them all. Share → ignitai. (Or lasso-select in Files and drag the lot across in Split View.)
One prompt — or your saved preset — applies across the batch.
The output is a single .xlsx, one row per record, with a source_file column tying every row back to its image. Fourteen receipt photos become one expense sheet you can subtotal by day.
Images that fail — a blurry frame, an accidental screenshot of the home screen — are listed separately, so you re-shoot or re-run only those.

An iPad-sized batch is a dozen or two images. If what you actually have is a folder of hundreds accumulated on a drive, that’s a desk job: the Mac batch workflow takes a folder in one drag — same pipeline, same presets, synced via iCloud.

A worked example: the dashboard with no export button

Your team’s analytics tool shows a table of last quarter’s numbers — forty rows of campaign, spend, clicks, conversions — and the export feature is on the plan above yours. You need it in Excel for the Monday deck:

In Safari on the iPad, scroll the table and take three overlapping screenshots.
Open ignitai in Split View next to Photos. Select the three screenshots and drag them across.
Prompt: “Extract the campaign table: campaign name, spend, clicks, conversions. One row per campaign. Ignore the sidebar, header, and chart.”
Output: Excel. Tap Extract. All three screenshots finish in seconds, merged into one sheet.
In the preview, delete the two duplicate boundary rows where the screenshots overlapped. Confirm the last campaign made it. Save to iCloud Drive.

Forty rows, zero retyping, and the paywall stayed unclicked. If the numbers should land in Apple’s spreadsheet instead, the PDF-to-Numbers flow on iPad is the same pattern with a different output pick.

The verification habit: make the sheet reproduce the picture

Extraction from an image deserves the same one-formula check as extraction from any scan: make the spreadsheet reproduce a number visible in the image. If the pictured table shows a total, =SUM() the column and compare — with the image sitting right beside the sheet in Split View, this takes ten seconds. If there’s no printed total, count: forty campaigns in the screenshot should be forty rows in the sheet.

Match: done — the sheet is trustworthy.
Mismatch: sort the column and scan the extremes. A glare spot turning an 8 into a 3 is findable in a minute, and the fix is re-shooting one image, not redoing the job.

A spreadsheet built from a picture is only as good as the one number you verified. Verify one.

A note on privacy

Pictures of documents are documents. A screenshot of an account page, a photographed pay statement, a texted medical bill — uploading them to a web “image to Excel” converter is uploading exactly the content the original would have protected. On-device extraction on iPadOS 26+ means the image is read on your iPad and the Excel file is written on your iPad; nothing leaves the device. The same caution that applies to bank statements on iPad applies to a picture of a bank statement — same data, different file extension.

Bottom line

A picture of a table on your iPad — screenshot, texted JPEG, synced photo — contains data iPadOS can read but can’t structure: Live Text gives you word soup, Photos has no table export, and web converters guess columns from pixels and lose. The workflow that works in 2026: drag or share the image into ignitai, name the table in one plain-English sentence if the image is busy, export .xlsx with the types already set, and verify one visible number with the image and the sheet side by side in Split View. On an M-series or A17 Pro iPad running iPadOS 26, the image never leaves the device. Use this path when the picture already exists; when the paper is in front of you and runs multiple pages, capture with the document scanner and take the scanned-PDF path instead.

Get ignitai on the App Store — free download, $19.99/mo unlocks unlimited extractions and batch mode after the 3-day trial.

FAQ

How do I convert a screenshot to Excel on iPad?

Take the screenshot, then share it into ignitai from Photos — or drag it across in Split View — add a one-sentence prompt if the screen contains more than the table, and choose .xlsx as the output. Screenshots are the easiest input there is: flat, evenly lit, machine-rendered text extracts essentially without error. For tables longer than one screen, share overlapping screenshots as a single job.

What image formats work for converting a picture to Excel on iPad?

HEIC, JPEG, PNG, and screenshots all work, shared or dragged directly from Photos, Mail, Messages, Safari, or Files. No conversion to PDF is needed first; the image is read as-is and the output is a standard .xlsx that opens in Excel, Numbers, or Google Sheets.

Does converting an image to Excel on iPad work offline?

On an M-series or A17 Pro iPad running iPadOS 26, yes — ignitai runs the extraction on-device using Apple’s Foundation Models framework, so it works in airplane mode and the image never leaves the iPad. Older iPads fall back to a hosted pipeline with documented zero retention, which requires a connection.