# Key Information Extraction on OCRBench, CCOCR, and Nanonets Suite (test)

Source: [Key Information Extraction on OCRBench, CCOCR, and Nanonets Suite (test)](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test>)

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

Data updated: 2026-03-17T08:00:09.959Z

Evaluation of Key Information Extraction (KIE) performance across multiple OCR-related datasets, including various versions of OCRBench, CCOCR, and Nanonets (v1.5/v2.0 test splits).

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 | Key Information Extraction |
| Dataset (including any stated split) | OCRBench, OCRBenchv2, CCOCR, Nanonets en zh v1.5 v2.0 (test) |
| Metric | Overall Mean Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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 Mean Score](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=Overall%20Mean%20Score>): Higher is better; unit: PERCENT.

- [OCRBench KIE Accuracy](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=OCRBench%20KIE%20Accuracy>): Higher is better; unit: PERCENT.

- [OCRBenchv2 KIE (en) Accuracy](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=OCRBenchv2%20KIE%20(en)%20Accuracy>): Higher is better; unit: PERCENT.

- [OCRBenchv2 KIE (zh) Accuracy](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=OCRBenchv2%20KIE%20(zh)%20Accuracy>): Higher is better; unit: PERCENT.

- [CCOCR KIE Accuracy](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=CCOCR%20KIE%20Accuracy>): Higher is better; unit: PERCENT.

- [Nanonets KIE F1 Score](<https://www.sota2.com/research/sota/key-information-extraction-on-ocrbench-ccocr-and-nanonets-suite-test.md?metric=Nanonets%20KIE%20F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Overall Mean Score | Result date | Report paper |
| --- | --- | --- | --- |
| Qianfan-OCR | 87.9 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Qwen3-VL-235B-A22B | 84.2 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Qwen3-4B-VL | 83.5 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Gemini-3.1-Pro | 79.2 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Seed-2.0 | 78 | Not recorded | 69b8e00a0636e5bf4afbea02 |
| Gemini-3-Pro | 77 | Not recorded | 69b8e00a0636e5bf4afbea02 |

### Qianfan-OCR

Overall Mean Score: 87.9. Result ID: 69b9095b47f34ce924a1597c.

[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: Not recorded. This record does not assert independent reproduction.

### Qwen3-VL-235B-A22B

Overall Mean Score: 84.2. Result ID: 69b9095b47f34ce924a1597e.

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

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

### Qwen3-4B-VL

Overall Mean Score: 83.5. Result ID: 69b9095b47f34ce924a1597d.

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

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

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

### Gemini-3.1-Pro

Overall Mean Score: 79.2. Result ID: 69b9095b47f34ce924a1597f.

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

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

### Seed-2.0

Overall Mean Score: 78. Result ID: 69b9095b47f34ce924a15981.

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

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

### Gemini-3-Pro

Overall Mean Score: 77. Result ID: 69b9095b47f34ce924a15980.

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