opendatalab/OmniDocBench

• mediumreadme#1

  Strengthen the README's opening sentence to emphasize its comprehensive evaluation role for MLLMs

  Why:

  CURRENT

  OmniDocBench is a benchmark for evaluating diverse document parsing in real-world scenarios, featuring the following characteristics:

  COPY-PASTE FIX

  Copy

  OmniDocBench is a comprehensive benchmark designed to evaluate multimodal large language models (MLLMs) and other document parsing models across diverse real-world scenarios, featuring:

  OpenDoneDismiss
• mediumcomparison#2

  Add a 'Comparison with Existing Benchmarks' section to the README

  Why:

  COPY-PASTE FIX

  Copy

  ## Comparison with Existing Benchmarks
  
  OmniDocBench distinguishes itself from existing document parsing benchmarks (e.g., DocVQA, FUNSD, SROIE, RVL-CDIP, CORD) through its comprehensive and integrated scope across multiple dimensions. Specifically, it offers:
  
  -   **Diverse Document Types:** Encompassing 10 document types, 5 layout types, and 5 language types across 1651 PDF pages, including academic papers, financial reports, newspapers, textbooks, and handwritten notes, providing a broader real-world evaluation context.
  -   **Rich Annotation Information:** Providing detailed localization, recognition results (text, LaTeX, HTML), and reading order for 28 block-level and 4 span-level elements, alongside extensive page and block-level attribute tags. This level of detail surpasses many existing datasets, enabling more granular and robust model evaluation.

  OpenDoneDismiss

Category GEO backends resolved for this scan: google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• CATEGORY QUERY

  What are the best benchmarks for evaluating document parsing models across diverse real-world scenarios?

  you: not recommended
  AI recommended (in order):

  1. DocVQA
  2. FUNSD
  3. SROIE
  4. RVL-CDIP
  5. CORD
  6. IIT-CDIP
  7. Tobacco800

  AI recommended 7 alternatives but never named opendatalab/OmniDocBench. This is the gap to close.

  Show full AI answer
• CATEGORY QUERY

  Where can I find a comprehensive dataset with rich structural annotations for various document types?

  you: not recommended
  AI recommended (in order):

  1. DocBank
  2. PubLayNet
  3. RVL-CDIP
  4. SROIE
  5. FUNSD
  6. XFUND

  AI recommended 6 alternatives but never named opendatalab/OmniDocBench. This is the gap to close.

  Show full AI answer

• Metadata completeness

  warn

  Suggestion:

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of opendatalab/OmniDocBench?

  pass
  ✓AI named opendatalab/OmniDocBench explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• If a team adopts opendatalab/OmniDocBench in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI named opendatalab/OmniDocBench explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• In one sentence, what problem does the repo opendatalab/OmniDocBench solve, and who is the primary audience?

  pass
  ✓AI named opendatalab/OmniDocBench explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?