Text recognition from photos and screenshots online
This assistant extracts text from photos and screenshots and provides it in a format convenient for copying and editing. It will transcribe text from a photo of a book page or document, recognize text on a screenshot of a conversation, interface, or presentation, preserve the structure of paragraphs and lists, transfer data from a photo of a table into a neat table in the chat, and, if necessary, translate the found text into another language or correct typos in it. To recognize text from a photo, simply attach the image and briefly write what format you need the result in. On BotHub, the assistant is available without a VPN or foreign card, is paid for in rubles, and chat history is saved, so it is easy to return to recognized fragments later.
Frequently Asked Questions
The assistant confidently recognizes printed text in photos of documents, book pages, signs, and screenshots. It also reads legible handwritten text, but accuracy depends on handwriting, lighting, and image clarity. Errors are possible on blurry images, with small font in the background, and with illegible notes, so it is better to retake such fragments closer and in even lighting.
Attach a screenshot to your message and write what result you need: continuous text, text with preserved paragraphs and lists, or a table. For example, you can upload a photo of a slide from a conference, ask to transcribe the text from it, and immediately shorten it to bullet points. The finished result can be copied from the chat and pasted into any editor or document.
If the image is clear and well-lit, the result usually almost matches the original, but it is still worth checking. Pay special attention to numbers, dates, names, details, and technical terms: an error in one character can change the meaning. If the source is important, for example, a contract or payment document, compare the recognized text with the image before using it.
Requests are deducted from a single balance in the internal currency CAPS. The cost depends on the number and size of uploaded images, the volume of recognized text, and additional actions, such as translation or editing the result. Several images in one message or a multi-page document consume more CAPS than a short screenshot with a couple of lines.
The assistant is pre-configured to work with text in images, so you don't need to explain the task and describe the response format every time. It preserves the structure of the original, does not add summaries or comments, and conveys the text as close as possible to what is shown in the image. In a regular chat, such a result often has to be achieved through follow-up requests.