docling/docs/concepts/confidence_scores.md

Introduction

Confidence scores were introduced in v2.34.0 to help users understand how well a conversion performed and guide decisions about post-processing workflows. They are available in the confidence field of the ConversionResult object returned by the document converter.

Purpose

Complex layouts, poor scan quality, or challenging formatting can lead to suboptimal document conversion results that may require additional attention or alternative conversion pipelines.

Confidence scores provide a quantitative assessment of document conversion quality. Each confidence report includes a numerical score (0.0 to 1.0) measuring conversion accuracy, and a quality grade (poor, fair, good, excellent) for quick interpretation.

Use cases for confidence scores include:

• Identify documents requiring manual review after the conversion
• Adjust conversion pipelines to the most appropriate for each document type
• Set confidence thresholds for unattended batch conversions
• Catch potential conversion issues early in your workflow

Concepts

Scores and grades

A confidence report contains scores and grades:

• Scores: Numerical values between 0.0 and 1.0, where higher values indicate better conversion quality
• Grades: Categorical quality assessments based on score thresholds:
  • POOR: Score < 0.5
  • FAIR: Score < 0.8
  • GOOD: Score < 0.9
  • EXCELLENT: Score ≥ 0.9

Types of scores

Each confidence report includes four component scores:

• layout_score: Overall quality of text content extraction
• ocr_score: Quality of OCR-extracted content
• parse_score: 10th percentile score of text cells (emphasizes problem areas)
• table_score: Table extraction quality (not yet implemented)

Summary scores

Two aggregate scores provide overall document quality assessment:

• mean_score: Average of the four component scores
• low_score: 5th percentile score (highlights worst-performing areas)

Both summary scores include corresponding mean_grade and low_grade fields for quick quality assessment.

Page-level vs document-level

Confidence scores are calculated at two levels:

• Page-level: Individual scores for each page, stored in the pages field
• Document-level: Overall scores for the entire document, calculated as averages of page-level scores and stored in fields equally named in the root ConfidenceReport

!!! note "Document-level scores"

For most use cases, users should safely focus on the document-level `mean_score` / `mean_grade` and `low_score` / `low_grade` fields to assess overall conversion quality.

Example