# Document Parsing on OlmOCR Bench v1.0 (test)

Source: [Document Parsing on OlmOCR Bench v1.0 (test)](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-17T08:00:09.928Z

Evaluation of Document Parsing performance on the test split of OlmOCR Bench version 1.0, covering specific sub-tasks like ArXiv math, table extraction, handling old scans, multi-column layouts, and long/tiny text.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Document Parsing |
| Dataset (including any stated split) | OlmOCR Bench v1.0 (test) |
| Metric | Overall Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Overall Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Overall%20Score>): Higher is better; unit: PERCENT.

- [ArXiv Math Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=ArXiv%20Math%20Score>): Higher is better; unit: PERCENT.

- [Old Scans Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Old%20Scans%20Score>): Higher is better; unit: PERCENT.

- [Table Extraction Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Table%20Extraction%20Score>): Higher is better; unit: PERCENT.

- [Header/Footer Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Header%2FFooter%20Score>): Higher is better; unit: PERCENT.

- [Multi-Column Layout Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Multi-Column%20Layout%20Score>): Higher is better; unit: PERCENT.

- [Long/Tiny Text Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Long%2FTiny%20Text%20Score>): Higher is better; unit: PERCENT.

- [Base Score](<https://www.sota2.com/research/sota/document-parsing-on-olmocr-bench-v1-0-test.md?metric=Base%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Overall Score | Result date | Report paper |
| --- | --- | --- | --- |
| PaddleOCR-VL | 80 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Qianfan-OCR | 79.8 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Qwen3-VL-4B | 79.2 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| PaddleOCR-VL-1.5 | 79.1 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Dotsocr | 79.1 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| DeepSeek-OCR | 77.2 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| MonkeyOCR | 75.8 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| MinerU2.5 | 75.2 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| PointsReader | 71.8 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Qwen3-VL-2B | 70.7 | Not recorded | 69b8e00a0636e5bf4afbea02 |

### PaddleOCR-VL

Overall Score: 80. Result ID: 69b9095b47f34ce924a15972.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3b70a8b01e76cc8da769.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"Pipeline"}. This record does not assert independent reproduction.

### Qianfan-OCR

Overall Score: 79.8. Result ID: 69b9095b47f34ce924a1597b.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Implementation](<https://github.com/baidubce/Qianfan-VL>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### Qwen3-VL-4B

Overall Score: 79.2. Result ID: 69b9095b47f34ce924a15979.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### PaddleOCR-VL-1.5

Overall Score: 79.1. Result ID: 69b9095b47f34ce924a15973.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3b70a8b01e76cc8da769.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"Pipeline"}. This record does not assert independent reproduction.

### Dotsocr

Overall Score: 79.1. Result ID: 69b9095b47f34ce924a15976.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### DeepSeek-OCR

Overall Score: 77.2. Result ID: 69b9095b47f34ce924a15978.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697ae9b8a2b28e761d87d127.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-OCR-2>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### MonkeyOCR

Overall Score: 75.8. Result ID: 69b9095b47f34ce924a15974.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"Pipeline"}. This record does not assert independent reproduction.

### MinerU2.5

Overall Score: 75.2. Result ID: 69b9095b47f34ce924a15975.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### PointsReader

Overall Score: 71.8. Result ID: 69b9095b47f34ce924a15977.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

### Qwen3-VL-2B

Overall Score: 70.7. Result ID: 69b9095b47f34ce924a1597a.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e00a0636e5bf4afbea02.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Architecture Paradigm":"End-to-end Model"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
