Extract text from any video — capture frames and read them with OCR. Your video stays in your browser.
No. Your video stays in your browser and is never uploaded. Only the individual frames you capture are sent for text recognition, which keeps your video private and avoids large uploads.
Capture frames where the text is large, sharp, and high-contrast. Blurry or fast-moving text is harder to read. The manual capture lets you pick exactly the clearest frames.
Manual lets you scrub to specific moments and grab just the frames with the text you want — best for accuracy. Auto-capture grabs a frame every few seconds across the whole video — convenient, but may produce duplicate text you can clean up with the dedupe option.
Text recognition currently runs in English. Clear printed or on-screen text works best; stylized or handwritten text is harder to read.
OCR needs readable text in the frame. If a frame has no clear text, small text, or heavy motion blur, nothing will be extracted. Try capturing a sharper frame where the text is larger.