# Long PDF Understanding on PaperPDF 50 English PDFs

Source: [Long PDF Understanding on PaperPDF 50 English PDFs](<https://www.sota2.com/research/sota/long-pdf-understanding-on-paperpdf-50-english-pdfs>)

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

Data updated: 2026-04-28T09:47:57.022Z

Evaluation of model performance on understanding long documents within the 'PaperPDF 50 English PDFs' dataset, measured using metrics such as ANLS, F1, Rouge, and GPT-Acc.

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 | Long PDF Understanding |
| Dataset (including any stated split) | PaperPDF 50 English PDFs |
| Metric | ANLS |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [ANLS](<https://www.sota2.com/research/sota/long-pdf-understanding-on-paperpdf-50-english-pdfs.md?metric=ANLS>): Higher is better; unit: PERCENT.

- [F1 Score](<https://www.sota2.com/research/sota/long-pdf-understanding-on-paperpdf-50-english-pdfs.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

- [ROUGE Score](<https://www.sota2.com/research/sota/long-pdf-understanding-on-paperpdf-50-english-pdfs.md?metric=ROUGE%20Score>): Higher is better; unit: PERCENT.

- [GPT Accuracy](<https://www.sota2.com/research/sota/long-pdf-understanding-on-paperpdf-50-english-pdfs.md?metric=GPT%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ANLS | Result date | Report paper |
| --- | --- | --- | --- |
| PDF-WuKong | 41.8 | Not recorded | 68fa76ad1e209131cea6e62a |
| Qwen | 36 | Not recorded | 68fa76ad1e209131cea6e62a |
| ChatGLM | 31.2 | Not recorded | 68fa76ad1e209131cea6e62a |
| Kimi | 28.5 | Not recorded | 68fa76ad1e209131cea6e62a |
| Gemini 1.5 pro | 26.6 | Not recorded | 68fa76ad1e209131cea6e62a |

### PDF-WuKong

ANLS: 41.8. Result ID: 69f0823a47f34ce924d6e188.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen

ANLS: 36. Result ID: 69f0823a47f34ce924d6e187.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### ChatGLM

ANLS: 31.2. Result ID: 69f0823a47f34ce924d6e186.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

Recorded evaluation configuration: {"Note":"Commercial product tested in Sep 2024"}. This record does not assert independent reproduction.

### Kimi

ANLS: 28.5. Result ID: 69f0823a47f34ce924d6e185.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Gemini 1.5 pro

ANLS: 26.6. Result ID: 69f0823a47f34ce924d6e184.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76ad1e209131cea6e62a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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