@transcriber_botFreemium
These bots read the text in a photo, a screenshot or a scanned page and send it back as editable text: a menu abroad, a page from a textbook, a contract you only have on paper.
| # | Bot | Price | Languages | Inline | Groups | Score |
|---|---|---|---|---|---|---|
| 1 | @transcriber_bot | Freemium · Transcription itself is free; the Telegram Stars tiers add features, not access. | EN | no | no | 81 |
| 2 | @imagetotext_bot | Free | EN | no | no | 73 |
| 3 | @TexifyBot | Freemium | EN · MULTI | no | no | 72 |
| 4 | @perebot | Free | RU · MULTI | no | no | 71 |
| 5 | @tesseract_ocr_bot | Free | EN · RU | no | no | 67 |
| 6 | @bacakubot | Freemium | EN · MULTI | no | no | 66 |
| 7 | @the_ocr_bot | Free | EN | no | no | 61 |
Every bot received the same four images: a sharp photo of a printed page in English, a phone shot of a Russian page taken at an angle, a screenshot with mixed Latin and Cyrillic text, and a handwritten note. We counted recognition errors per hundred characters, checked whether paragraph breaks survived, measured the reply time and noted the free limit and whether the bot keeps the uploaded image.
Accuracy on printed text is table stakes; the differences show up on photos taken at an angle, on low light and on mixed alphabets. A bot that supports over 100 languages, like Texify, usually detects the language itself, while a single-language bot may read Cyrillic as garbled Latin. Check whether the result keeps line breaks and paragraphs: for a table or a receipt that matters more than a single wrong letter.
Most bots accept a photo sent as an image; sending it as a file keeps the original resolution and gives visibly better results on small print. Multi-page PDFs are handled by only a few bots and are usually where the paid tier starts. Handwriting recognition remains hit-and-miss in every bot we tested: a neat note works, a shopping list does not.
The image is uploaded to the developer's server before recognition, so avoid documents with passport data or bank details unless the bot states how long it keeps files. Free tiers are typically limited by the number of images per day or per month rather than by quality; bots built on the open-source Tesseract engine have no limits but lag behind on difficult photos.
Texify advertises over 100 languages and detects the language on its own, and in our test Texify, Perebot and the Tesseract bot also read Cyrillic. Image To Text is tuned for English, and the other bots publish no list of the languages they recognise, so try a page in your language before relying on one.
Image To Text and the Tesseract bot are free with no paid plan and no usage limit mentioned; Texify and OCR Bot keep the best quality for paid plans after a free quota.
Perebot and OCR Bot (@bacakubot) do recognition and translation in one step; the others return text only, which you can paste into any translator bot.
Perspective, shadows and compression matter. Shoot straight above the page in daylight and send the image as a file, not as a photo, to keep full resolution.
![Image To Text [OCR]](/media/imagetotext_bot-96x96.jpg)
