Prepare one image, choose its text language and recognition method, then check the extracted text and download the version you need. Follow the real example below, then adapt it to your own source. Open Image to Text.
Upload a readable text image
Open Image to Text and upload the clear English text image shown in the comparison. Frame the letters with the selection area; keep the source upright and avoid clipping an edge.
Upload an image or open the sample. Use the selection handles to crop the text area, and use the available enhancement controls when brightness, contrast or the background makes the letters difficult to read. A clear, upright source with readable characters is a better starting point than a tiny or blurred screenshot.
Maximum input size: 25 MB. One image per run. Image dimensions and available memory can impose additional limits.
Choose recognition settings
Choose English for this source, change the initial Future method to Classic, and leave the recognition area on the visible text. The example uses the settings shown below.
Choose the language that matches the text. Future uses AI recognition. Classic uses the established OCR engine and provides single-column or multiple-column layout settings. Choose the method and layout for the source, then select Start and complete the verification challenge.
The example in this guide
Verified in the i2TEXT browser workflow on November 11, 2026.
- Input
- clear-text.png
- Output
- english-ocr-page-1.txt
- Saved result
- 59 bytes
Recognize and review the words
Select Start and complete the verification challenge. When recognition finishes, compare the text pane with the image, especially digits, punctuation and line breaks. You can correct the editable text before saving.
OCR can confuse similar letters, miss small text or change reading order. Review the result against the image before relying on it. An animated or multipage source is treated as a still image rather than a full document batch.
Save the text you reviewed
Choose the TXT download to save english-ocr-page-1.txt. TXT and Word use the current text. The comparison shows the actual TXT saved during this run.
TXT and Word downloads use the current edited text. With Future, Simple PDF also uses current text. With Classic, Searchable PDF keeps the prepared image and the original recognized text layer, so later text edits do not rewrite that PDF layer. Pick the format that matches what you want to preserve.
clear-text.pngenglish-ocr-page-1.txt
Common questions
Do corrections change a Classic searchable PDF?
TXT and Word downloads use the current edited text. With Future, Simple PDF also uses current text. With Classic, Searchable PDF keeps the prepared image and the original recognized text layer, so later text edits do not rewrite that PDF layer. Pick the format that matches what you want to preserve.
What should I do with a difficult source image?
Upload an image or open the sample. Use the selection handles to crop the text area, and use the available enhancement controls when brightness, contrast or the background makes the letters difficult to read. A clear, upright source with readable characters is a better starting point than a tiny or blurred screenshot.
Can I use the text in another tool?
Translate can process the current extracted text up to 25,000 characters. Smart Edit provides formatting and optional AI text actions; an AI action accepts up to 600 words and 12,000 characters. Review any changed wording. Continue to can send the current text to a compatible i2TEXT tool for a separate next step.
Try Image to Text with your own source.
Use your own source and review the result before sharing.
