What Is the Best Table Recognition Model?
Compare TeleOCR's table recognition results, understand TEDS and TEDS-Structure, and learn why benchmark scores need testing on your own tables.
AI recognition / Free to use
Convert a JPG, PNG, photo, or screenshot into an editable text table with AI OCR, then export it to Excel, CSV, Markdown, HTML, or JSON.
Click to upload or drag and drop
JPG, PNG, WebP, GIF or BMP (max 5MB, 8192x8192px; animations use the first frame)
Your table editor is ready
Upload a table image above. You can edit every recognized cell before exporting.
No table to export yet
Upload an image and review the recognized table. Your Markdown output will appear here automatically.
Workflow
Choose a clear screenshot, scan, or photo. If the image contains extra content, use Image Cropper to crop it around the table. Keep the text upright so rows and columns are easier to recognize.
Fast (HunyuanOCR-1.5) is the default. For better results, select Accurate (TeleOCR) before uploading, or switch to Accurate and click Recognize again if the initial result is poor. The image is sent to the selected recognition service. Review and correct text, numbers, or merged cells.
Choose Excel, CSV, Markdown, HTML, JSON, or another available format. Copy the text output or download the file after checking the result.
Documentation
Compare TeleOCR's table recognition results, understand TEDS and TEDS-Structure, and learn why benchmark scores need testing on your own tables.
See real table images and OCR results to learn when local recognition is enough and when to choose AI Image to Table for complex layouts.
How imgtotable uses HunyuanOCR to generate HTML from a table image, then rebuilds editable cells and merged headers for spreadsheet export.
Choose AI or local table recognition, check rows and values, and export a table screenshot to Excel, CSV, or Markdown.
Explore
Recognize and edit table grids on your device without uploading the selected image.
Turn document images into editable Markdown with a source editor and rendered preview.
Extract editable plain text with browser-based OCR while the image stays on your device.
Help
Yes. You can recognize table images, edit the results, and export your tables for free in both Fast and Accurate modes.
Upload PNG, JPEG, WebP, GIF, or BMP images up to 5 MB and 8192 × 8192 pixels. Animated images use the first frame. A sharp, tightly cropped table usually gives a clearer result.
Fast uses HunyuanOCR-1.5 and is the default. Accurate uses TeleOCR and is a useful option for scanned tables or complex layouts. Results depend on the image, so check the table before exporting. You can choose the other mode and run recognition again.
This AI tool uploads the image for server-side recognition. Editing and final export happen in your browser. If you want recognition without uploading your image, use Local Image to Table in the Related tools section.
Yes. Review and edit the recognized cells, then select Excel and download the file. You can also export CSV, Markdown, HTML, JSON, LaTeX, XML, ASCII, MediaWiki, or YAML. Check numbers, headings, and merged cells against the source image first.