Doc2X Docs

English

简体中文

English

简体中文

Theme

Doc2X Common Questions - FAQ

Doc2X is a powerful document parsing and translation tool that supports recognition, conversion, and translation of multiple document formats. This guide will help you quickly understand and use Doc2X's various features.

Translation - Points Rules

Doc2X has updated the translation points rules for web and client versions. For details, please refer to Doc2X Points Rules.

Currently Supported Features

Core Recognition Features

  • ✅ Multi-element Recognition: Supports recognition of text, formulas, tables, and images
  • ✅ Multi-column Recognition: Supports recognition of multi-column documents and restores correct reading order
  • ✅ Multi-language Support: Supports Chinese (Simplified/Traditional), English, Western European languages, Japanese, Korean, etc.
    • More languages are being supported

Advanced Features (In Development)

The following features are supported but their effectiveness is continuously being optimized:

Current Limitations

  • ❌ Does not support extra-long/extra-wide images: Need to manually split into normal page sizes
  • ❌ Does not support documents with excessive blank margins: Need to manually crop blank margins
  • ❌ Does not support rotated PDFs: Please manually rotate PDF to correct orientation before recognition

Processing Speed and Concurrency

Processing Speed

  • Web and API: Average speed for single PDF is approximately 10 pages/second
  • Actual Speed: Depends on document complexity
  • API Acceleration: Contact us for faster processing speeds

Concurrency Limits

  • Default API Concurrency: 10 PDFs processed simultaneously
  • Increase Concurrency: Contact WeChat customer service if you need higher concurrency

Large Batch Data Processing

Enterprise Services

  • Large Volume Processing: Contact us for discount pricing on large PDF processing volumes
  • Served Clients: Multiple well-known large language model companies, financial and educational institutions
  • Infrastructure: Self-built hundreds of GPU compute pools and multi-data center redundancy ensuring stability
  • Processing Capacity: Can process tens of millions of document pages daily, with cumulative processing of billions of pages

Why Choose Doc2X

Core Advantages

  1. Leading Formula Recognition: Similar products on the market perform poorly in formula recognition (especially inline and complex formulas), while Doc2X is at the leading level
  2. Excellent Table Recognition: Supports advanced features like recognizing images within tables and cross-page table merging
  3. Precise Multi-column Recognition: Outstanding performance in restoring reading order for multi-column documents
  4. Wide Adaptation Range: Supports various document types including financial research reports, academic papers, educational materials, patents, etc.

Data Security

  • Web Storage: Valid for 30 days (including image hosting)
  • API Storage: Expires after 24 hours
  • Automatic Deletion: Automatically deleted after expiration, please use with confidence

PPT Generator FAQ

For illustrated steps, see the AI PPT tutorial. For questions you have already reviewed, see Teaching Slides from Reviewed Questions.

1. What Is the Standard PPT Workflow?

Open Utilities → Magic Pic → AI Generate PPT. There are two main paths:

  • Smart Layout (智能排版): Topic or parsed PDFs → Generate presentation → Edit text, formulas, images and shapes → Export PPTX.
  • Creative Images (创意生图): Topic or parsed PDFs → Generate and adopt outline → Edit outline → Page design → Generate all images → Preview and export.

Creative Images can export an image deck directly. Run Edit text and images (编辑文字与配图) when you want to edit text inside the images or export an editable deck.

2. How Can Outlines and Prompts Be Generated When Settings Focus on Image AI?

The default creation page uses its own AI model selection for topic/document generation. It does not rely on a third-party Key in image settings. Doc2X image generation uses account points without a Key; Custom service exposes third-party model, Key and Base URL fields.

The older One sentence / Paste outline workflow and AI outline refinement still require a usable custom text-model configuration. Without one, return to Generate from reference documents (参考文档生成). You can also edit an existing outline manually and select Generate page design. The older workflow's configuration requirements do not apply to all PPT features.

3. Which Image Providers, Ratios and Sizes Are Available?

Settings offer Doc2X image generation and Custom service. The former uses the currently available platform image model; the latter supports OpenAI or Gemini with your model and service details.

On the default creation page, Smart Layout uses 16:9, while Creative Images offers 16:9 / 4:3. Custom image sizes depend on the provider and model, so check the available choices in Settings. The presentation ratio and the image service's output pixel ratio may differ.

4. Why Does the Ratio or Size Change When I Switch Providers?

Services support different sizes. Switching a custom Provider adjusts the saved size to a supported value; the older topic/outline entry may also adjust its ratio. Check size, quality and cost notices again before generation. Each Provider keeps its own Key, model and address.

5. What Does Concurrency Mean?

It is the number of image requests processed at once. The setting allows 1–8, but Doc2X image generation processes at most 2 slides concurrently per batch; custom services allow up to 8. Lower the value if you encounter rate limits. A higher value does not guarantee faster results.

6. Why Does Generate All Images Replace Old Images and Parsing Records?

It regenerates slides with design instructions, replacing their images and clearing old drawing-board associations so that new images do not reuse edits from old images. It is not a failed-slides-only retry. Export any version you want to keep first.

A slide's Redesign (重新设计) action only changes design instructions. The standard preview does not provide a single-slide image regeneration button.

7. Why Should I Keep the Page Open During Image Generation?

The page tracks progress and saves completed images. Each completed image is saved automatically; Cancel generation (取消生成) stops unfinished work and retains completed images.

Completed results from the default topic/document AI workflow can also be recovered from cloud-project entries. If a generated result cannot be opened, try Reopen result (重新打开结果) before starting generation again.

8. Why Is Export Disabled, or Why Does It Include Only Some Slides?

The current button is Export PPTX (导出 PPTX). It may be unavailable when there are no usable slides or while batch parsing is running. Requirements differ by format:

  • Image PowerPoint (图片版 PowerPoint): Uses generated slide images; no prior text parsing is required.
  • Editable PowerPoint (可编辑 PowerPoint): Includes slides with available drawing-board parsing records only.

Check the ready-page count in the export dialog. For a full editable deck, run Edit text and images and confirm that every slide has been parsed successfully.

9. How Do Editable and Image PPTX Differ?

For Creative Images, Editable PowerPoint uses parsed text and illustrations for further editing. Image PowerPoint keeps each slide as a single image. Both produce .pptx files.

Smart Layout already provides independently editable text, formulas, images and shapes. It can Export PPTX directly without an image-parsing step.

10. Why Does Batch Parsing Ask Me to Sign In?

The current action is called Edit text and images (编辑文字与配图). Creating new parsing records requires sign-in; existing usable records can be reused. After parsing, the workspace tries to open an available slide. You can also click Edit this slide (编辑此页) on a preview card.

If local drawing-board data is missing, opening or exporting an existing record may also require sign-in to retrieve cloud data.

11. Why Do Preview and Export Change After Drawing-Board Edits?

Creative Images preview and export prefer the latest available drawing-board edits, including changed text and images. Check the preview again after editing. Regenerating a slide image clears its old association, so it must be parsed again before further drawing-board editing.

12. Where Are Projects and Images Stored, and What Happens When I Clear Cache?

Projects, generated images and later edits are stored in the current browser or client's local cache. Drawing-board history uses a separate cache. Clearing PPT project cache deletes local projects and images.

The project list also includes completed cloud AI presentation tasks for the current account. While a task remains available, you can reopen its original generated result. Later local edits do not automatically sync to other devices. Download the final PPTX before clearing cache or switching devices. Cloud records can expire and are not a permanent backup.

13. What Does an Expired Drawing-Board Record Mean?

The slide's linked parsing record is no longer available. Return to Preview and export, run Edit text and images, and wait for a new record before opening or exporting the slide. This message does not by itself require regenerating the entire image deck.

14. Can I Customize Styles and AI Prompts?

Yes. The default creation page offers Editorial, Research and Business styles. In Creative Images, you can also edit each slide's outline, Page narrative and layout, and design instructions.

Style customization and AI prompt customization in Settings affect the traditional styles and outline/image-design prompts. Empty fields use defaults. These changes do not rewrite existing slides or serve as a separate configuration form for the default document AI model.