README.md

metadata

license: other
license_name: dharmaocr-benchmark-license
license_link: LICENSE
pretty_name: DharmaOCR-Benchmark
language:
  - pt
task_categories:
  - image-text-to-text
tags:
  - ocr
  - benchmark
  - document-understanding
  - brazilian-portuguese
  - text-recognition
  - handwriting-recognition
  - legal-documents
size_categories:
  - n<1K
dataset_info:
  features:
    - name: id
      dtype: int64
    - name: image
      dtype: image
    - name: image_base64
      dtype: string
    - name: assistant
      dtype: string
    - name: assistant_without_json
      dtype: string
  splits:
    - name: test
      num_bytes: 2534593858
      num_examples: 496
  download_size: 2532700883
  dataset_size: 2534593858
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*

DharmaOCR-Benchmark

Overview

DharmaOCR-Benchmark is a 496-instance evaluation suite for OCR models focused on Brazilian Portuguese documents. It covers printed text, handwritten text, and legal/administrative documents — domains underrepresented in existing benchmarks like OCRBench and olmOCR-Bench.

This benchmark evaluates not only transcription quality, but also text degeneration rate and unit inference cost as first-class metrics.

Released alongside the DharmaOCR family of models. For the full methodology and analysis, see our paper: DharmaOCR: Specialized Small Language Models for Structured OCR that Outperform Open-Source and Commercial Baselines.

Why this benchmark?

Existing OCR benchmarks do not reliably predict performance on Brazilian Portuguese documents. Language-specific orthography, domain vocabulary, and document formatting shift error profiles and amplify text degeneration in ways that general-purpose benchmarks fail to capture.

DharmaOCR-Benchmark fills this gap with a focused, reproducible evaluation protocol.

Dataset Composition

Subset	Samples	Description
ESTER-Pt	363	Printed text recognition in Brazilian Portuguese
Legal	83	Legal and administrative documents (publicly sourced, fully human-audited)
BRESSAY	50	Handwritten text recognition in Brazilian Portuguese
Total	496

⚠️ This benchmark was not used for training, model selection, DPO pair construction, or quantization calibration of any DharmaOCR model.

Evaluation Protocol

Score

DharmaOCR-Benchmark Score = (LevenshteinRatio + BLEU) / 2

Component	What it captures
LevenshteinRatio	Character-level fidelity (misspellings, missing accents, punctuation)
BLEU	N-gram sequence preservation (reorderings, dropped spans)

Additional Metrics

• Text degeneration rate (%): Requests that hit the output-token limit and exhibit repeated text spans (n-gram criterion). A critical operational metric — degenerate requests inflate cost and reduce throughput system-wide.
• Unit cost per page: Enables fair comparison between self-hosted models and commercial APIs.

Inference Setup

Parameter	Value
GPU	NVIDIA L40S (48GB GDDR6)
Instance	AWS g6e.2xlarge
Engine	vLLM
Max output tokens	8,192
Temperature	0

🏆 Benchmark Results

Model	Score ↑	Degeneration Rate (%) ↓	Time/Page (s) ↓
🥇 DharmaOCR Full (7B, ours)	0.925	0.40	2.132
🥈 DharmaOCR Lite (3B, ours)	0.911	0.20 ✨	1.464
Commercial APIs
Claude Opus 4.6	0.833	—	—
Gemini 3.1 Pro	0.820	—	—
GPT-5.4	0.750	—	—
Google Vision	0.686	—	—
Google Document AI	0.640	—	—
GPT-4o	0.635	—	—
Amazon Textract	0.618	—	—
Mistral OCR 3	0.574	—	—
Open-Source Models
Qwen2.5-VL-7B-Instruct	0.839	2.42	3.101
Qwen3-VL-8B	0.829	5.65	7.250
olmOCR-2-7B	0.823	1.41	4.306
Nanonets-OCR2-3B	0.791	2.62	1.911
Dots OCR	0.738	6.85	2.526
GLM-OCR	0.710	11.69	1.480
Qwen3-VL-2B-Instruct	0.623	11.69	3.566
Qwen2.5-VL-3B-Instruct	0.549	0.60	1.500
gemma-3-4b-it	0.214	33.96	2.182
DeepSeek-OCR	0.196	21.98	1.213

_{Score = (LevenshteinRatio + BLEU) / 2. Time/page on NVIDIA L40S. ✨ = lowest degeneration rate across all models.}

Usage

from datasets import load_dataset

dataset = load_dataset("dharma-ai/DharmaOCR-Benchmark")

Citation

@misc{cardoso2026dharmaocrspecializedsmalllanguage,
      title={DharmaOCR: Specialized Small Language Models for Structured OCR that outperform Open-Source and Commercial Baselines}, 
      author={Gabriel Pimenta de Freitas Cardoso and Caio Lucas da Silva Chacon and Jonas Felipe da Fonseca Oliveira and Paulo Henrique de Medeiros Araujo},
      year={2026},
      eprint={2604.14314},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2604.14314}, 
}

Contact

For technical questions, benchmark usage, research inquiries, or paper-related discussions:

gabriel.pimenta@dharma-ai.com.br