Image to Text Converter

Pull text straight out of screenshots, photos, scanned pages, or burned-in captions. Choose Fast mode for quick images or Accurate mode for harder images; Accurate mode downloads about 31 MB once in your browser.

Recognition mode

Runs entirely in your browser — your image is never uploaded.

This reads all visible text in the image, so it may include handles, timestamps, watermarks, or other text that is not a caption.

100% private: files are processed in your browser and never uploaded.

How it works

  1. Step 1

    Drop an image into the box above, or choose a PNG, JPG, or WebP file.

  2. Step 2

    Choose Fast mode for a smaller download, or Accurate mode for a larger download (about 31 MB) when precision matters.

  3. Step 3

    Copy the extracted text, or download it as a .txt — nothing is ever uploaded.

Want to do more with your subtitles?

Need to edit subtitles against the video, transcribe, translate, or burn captions into the video itself? Captioner is a full subtitle editor with AI transcription in 100+ languages, translation, and styled caption rendering.

Open in the Captioner editor

Frequently asked questions

How do I extract text from an image?
Choose Fast mode for quick screenshots or Accurate mode for harder images, then drop a PNG, JPG, or WebP into the box above. The tool reads the words in your browser and lists them as text you can copy or download as a .txt. Nothing is uploaded to a server.
Is my image uploaded anywhere?
No. Unlike most image-to-text tools, the recognition runs entirely in your browser using on-device OCR. Your image never leaves your computer or phone — it is not sent to Captioner or anyone else.
What image formats can I use?
PNG, JPG, and WebP. Screenshots and photos both work; the clearer and higher-contrast the text, the better the result.
Can it extract text from a screenshot or a photo?
Yes — screenshots, photos of signs or documents, slides, and social posts all work. High-contrast, roughly horizontal text reads best; very stylized, rotated, or handwritten text may come out imperfect.
Can it pull captions or subtitles out of an image?
It reads burned-in captions along with everything else visible in the image. Note that it extracts all visible text, so handles, timestamps, and watermarks can show up too — caption-only filtering is on the way. To create or edit real subtitle files for a video, use the Captioner editor.
What languages does it support?
It recognizes English and many other Latin-script languages, plus Chinese and several more. Accuracy is strongest for clear printed English; other scripts and low-quality images vary. (Japanese is not supported in this version.)
Is it really free?
Yes. No signup, no watermark, no upload — extract text from as many images as you like, free.

See all free tools or check Captioner pricing.