Study guides · Updated October 2026
Image summary: how to turn a photo of a page into notes you can revise from
An image summary is two jobs in one. It gets the words out of a photo or screenshot the way an image to text converter does, then shortens them into something you can actually study: the headings, the points that matter, and what the chart on the page is saying. This guide covers how that differs from plain OCR, how to ask for a photo summary rather than a transcript, and why a badly lit photo comes back half-read.
What you upload
A full textbook page, a lecture slide, or a photo of your own handwriting.
What comes back
- One or two lines on what the page is
- The text it could read, extracted
- The main ideas as bullet points
- What any chart or diagram shows
Maths is transcribed in LaTeX, not solved — unless you ask.
Free to try, no subscription needed for the default model.
Image summary vs OCR: where the two part company
An image to text converter finishes its job the moment the characters are recognised. You get a wall of raw text in exactly the order it appeared on the page, running headers, page numbers and footnote markers included, and then it is your problem to read. That is useful when you need the words verbatim — a quotation, a reference, a code snippet.
A summary carries on from there. It keeps the ideas and drops the clutter, so a two-page spread becomes something you can get through in a minute. The trade is precision for speed: a summary is a compression of the page, so anything you need word-for-word should be asked for as text rather than trusted to the short version.
Both come out of the same upload on Image Summary — the prompt box decides which, and the next section is how.
One upload, three things you can ask for
You do not switch tools to switch modes. The same cropped image behaves three ways depending on what you type, or do not type, in the prompt box:
A photo summary
Leave the prompt box empty.
The short version — the headings, the points that matter, nothing else.
An image to text transcript
Ask for all the readable text.
The words in the order they appeared on the page, which is the OCR output on its own.
A diagram explanation
Ask what the chart or diagram means.
The relationship it shows, described in words, rather than a reading of the axis labels.
After the first answer you can keep typing, and the follow-up box carries the conversation rather than the picture. That makes it good at reshaping what has already been said — a shorter version, a term explained, the bullets turned into test questions — and it means anything needing fresh detail from the image belongs in the first request. If you think you might want the exact wording later, ask for the transcript up front, so the words are in the conversation for every follow-up after it.
How to summarize an image, step by step
- Get the image in. Drag a file onto the page, pick one from your device, or use Take photo to shoot it with the rear camera there and then. PNG, JPEG and WEBP all work.
- Crop to what matters. Drag the corners around the paragraph, the table or the diagram you care about. This runs in your browser, and only the area you keep is uploaded.
- Say what you want, or do not. Empty box gives you the summary. Ask for the text and you get the transcript. Ask about the chart and you get the chart explained.
- Read it, then dig in. The answer streams in as it is written, and you can keep asking about the same image afterwards.
Screenshot to text is the easy case; a photo is not
A screenshot summary is close to exact, because a screenshot has no shadow, no angle and no blur — the characters are already crisp pixels. Slides, PDF pages and web pages all behave this way. A photo of a physical page is where results vary, and almost always for one of four reasons:
Fixing the photo beats fixing the prompt. No amount of instruction recovers text the camera never captured.
What it deliberately will not do
It is a summarizer, not a solver. Point it at a page of equations and it will transcribe the maths in LaTeX and tell you what the exercise is asking, but it stops short of working through the answer. That is on purpose: a summary of a worksheet is a study aid, whereas a worked solution is a different request with different expectations.
When the worked steps are what you are after, use Ask Questions, which reads the problem out of a photo or from typed text and shows the method, or Video Answers for the same thing as a narrated lesson. You can also simply ask Image Summary to solve it once the summary is on screen — it will, if you say so.
Privacy: what leaves your device
Cropping happens in your browser, so the part of the photo you cropped away is never transmitted at all — useful when the page you want is on a desk covered in other things. Summarizing itself is a stateless request: the image is held only for as long as it takes to produce your answer, then discarded rather than filed against your account, and uploads are not used to train public models. The privacy policy has the detail.
Image summary FAQ
Is an image summary the same as OCR?
OCR is the first half of it. OCR reads the characters out of a picture and hands them back in the order they appeared, page numbers and footnotes included. An image summary does that step and then explains what the text means — the main points, what a chart is showing, how a labelled diagram fits together. Ask for the raw extracted text and you get the OCR output on its own.
What is the difference between an image summary and a photo summary?
Nothing, as far as the tool is concerned. A photo you took and a screenshot you saved go through the same steps. People search for both phrases, which is why you see both, but it is one feature.
Which image formats work?
PNG, JPEG and WEBP. Screenshots of slides, PDF pages and web pages work, as do photos of printed books and handwritten notes. Crop to the section you need for the cleanest result.
Does it read handwriting?
Usually, if the handwriting is reasonably neat and the photo is well lit. Printed text is more reliable. Faint pencil, cursive and shadowed pages are the cases most likely to come back with gaps, so shoot in daylight and crop close.
Will it solve my maths homework from a photo?
Not by default. It transcribes the maths it can see using LaTeX and explains what the page is asking, but it will not work through the solution unless you ask it to. If worked steps are what you want, use Ask Questions, which takes a photo of the page or typed text, or ask explicitly to solve after the summary appears.
Can it explain a chart or a table?
Yes. It describes what the chart or table shows and lists the takeaways in plain language rather than only transcribing the numbers.
Can I ask follow-up questions about the same image?
Yes, and the follow-up box carries the conversation rather than the image itself. So asking for a shorter version, for a term to be explained or for the bullets as test questions all work from what has already been written. Anything that needs fresh detail from the picture — the exact wording of a paragraph the summary skipped — should go in the first request instead, or ask for the full transcript up front so those words are in the conversation.
What happens to the photo I upload?
Cropping happens in your browser, so only the area you approved is sent — the rest of the photo never leaves your device. Summarizing runs as a stateless request: the image is held for as long as it takes to produce the answer, then discarded rather than filed against your account. StudyHobby does not use uploads to train public models.
Is it free?
Yes. Converting an image to text and summarizing it is free to try without a subscription, and one StudyHobby login covers this and the other tools. Only the advanced models in the model picker need Pro.
Can I get the summary in a language other than English?
Yes. Pick your language with the language button at the top of the page and the summary, the headings and the bullet points all come back in it. The tool page itself is also published in Hindi, Urdu, Russian and Simplified Chinese.
Related image, OCR & summary guides
- AI image summarizerTurn a page of notes into bullet points.
- Image to text converterGet the words out of a photo or screenshot.
- Picture answerAsk a question about a photo you took.
- Photo mathPoint a camera at a maths problem.
- Screenshot notes from videoCapture a slide and turn it into notes.
Tools: Image Summary · Ask Questions · Chat Notes