run OCR for images and PDFs with structured Markdown, page JSON, table blocks, formulas, and layout-aware review in one workspace.
Best for scanned reports, statements, tables, slides, and dense document screenshots.
Drag and drop your file here, or click to select
Supports JPEG, PNG, GIF, WebP, BMP, TIFF, HEIC, PDF
Maximum file size: 20MB
Batch upload: up to 20 files
Paste an image or image URL from clipboard
This page is positioned for users looking for PaddleOCR-VL style document parsing: upload files, preserve layout hints, review recognized regions, and export usable Markdown or JSON.
Use formatted mode for headings, paragraphs, tables, formulas, and page-level blocks.
Normalized layout regions can be highlighted against the source preview when bounding boxes are available.
Correct OCR blocks after recognition while keeping the original result available.
Move from the web workspace to API keys, usage logs, credits, and operational monitoring.
Choose this page when the goal is structured document output, not only a flat text dump.
| Capability | Image to Text Web and API | Plain OCR Text only | Custom stack Build yourself |
|---|---|---|---|
| Output | |||
| Markdown and plain text | Text only | Depends | |
| Page JSON with block types | Depends | ||
| Tables and formulas | Formatted mode | Requires extra code | |
| Workflow | |||
| Source preview and review | Custom UI | ||
| History and deletion | Custom storage | ||
| API usage logs | Custom metering | ||
Formatted OCR aims to return content that can be reviewed, copied, exported, or integrated.
# Quarterly Report | Revenue increased 12% QoQ while support tickets fell 8%.
| Region | Revenue | Margin | with rows preserved for spreadsheet-friendly export.
pages[0].blocks includes type, text, confidence, page, and bbox fields when available.
Practical notes for users evaluating layout OCR.
Upload a document, review the structured output, and continue with API access when you need automation.