Lyrics from image (OCR)
Extract lyrics or any text from a screenshot, photo, or scanned booklet. AI vision OCR preserves line breaks and repeated lines exactly as written — paste the result straight into the LRC maker or lyric editor.
AI1 credit
Image OCR isn't enabled on this server yet.
Questions answered
Frequently asked questions
Any image file your browser can hand over — screenshots, phone photos, scanned booklet pages — in PNG, JPG, WEBP or any other image format. The server checks that the upload is an image rather than policing the extension. The size ceiling is set per server and shown on the upload box; it is 15 MB by default. Cropping to just the lyric text gives the cleanest result.
One credit per successful extraction, and you need to be signed in. Nothing is charged when it fails: the credit is only deducted once text has actually come back, so a blurry photo, an outage or an image with no readable text all cost you zero.
Yes. The extraction is asked for every visible line exactly as written, in the original order, including repeated chorus lines — nothing is deduplicated or tidied up. Line breaks come through as they appear in the picture, so the text drops straight into an LRC editor.
No. The picture is uploaded to our servers, read by an AI vision model, and the extracted text is sent back — the result card tells you whether that ran on our own hardware or a hosted model. Nothing is done locally in your browser.
Plain text you can copy or download as a .txt file. It is the words only — there are no timestamps, so take the result to the Manual LRC Maker or the Lyrics Aligner to turn it into a synced LRC.