# Optical Character Recognition on Bentham (test)

Source: [Optical Character Recognition on Bentham (test)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test>)

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

Data updated: 2026-04-21T08:35:53.750Z

Evaluation of Optical Character Recognition performance on the test split of the Bentham dataset, focusing on Character Error Rate (CER) and operational costs/token usage.

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 | Optical Character Recognition |
| Dataset (including any stated split) | Bentham (test) |
| Metric | CER |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 19 of 19 |
| 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

- [CER](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=CER>): Lower is better; unit: PERCENT.

- [Total Cost ($/1k docs)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Total%20Cost%20(%24%2F1k%20docs)>): Lower is better; unit: SCALAR.

- [OCR Cost ($/1k docs)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=OCR%20Cost%20(%24%2F1k%20docs)>): Lower is better; unit: SCALAR.

- [Input Cost ($/1k docs)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Input%20Cost%20(%24%2F1k%20docs)>): Lower is better; unit: SCALAR.

- [Output Cost ($/1k docs)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Output%20Cost%20(%24%2F1k%20docs)>): Lower is better; unit: SCALAR.

- [Tokens Image (/doc)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Tokens%20Image%20(%2Fdoc)>): Higher is better; unit: SCALAR.

- [Input Tokens (/doc)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Input%20Tokens%20(%2Fdoc)>): Higher is better; unit: SCALAR.

- [Output Tokens (/doc)](<https://www.sota2.com/research/sota/optical-character-recognition-on-bentham-test.md?metric=Output%20Tokens%20(%2Fdoc)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | CER | Result date | Report paper |
| --- | --- | --- | --- |
| OCR+PAGEN | 8.48 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+PAGE1 | 8.56 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+IMAGES | 8.67 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+PAGEN | 8.87 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+IMAGES | 9.35 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR | 9.54 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES | 9.74 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES | 9.87 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES-ALL-AT-ONCE | 9.88 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR | 9.97 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+IMAGES | 9.98 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES-ALL-AT-ONCE | 10.18 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR | 10.75 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+PAGE1 | 10.89 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+PAGE1 | 10.95 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR+PAGEN | 11.02 | Not recorded | 68fa772f1e209131cea71aa0 |
| OCR | 11.18 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES | 15.6 | Not recorded | 68fa772f1e209131cea71aa0 |
| IMAGES-ALL-AT-ONCE | 25.06 | Not recorded | 68fa772f1e209131cea71aa0 |

### OCR+PAGEN

CER: 8.48. Result ID: 69e736e247f34ce924cd7305.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### OCR+PAGE1

CER: 8.56. Result ID: 69e736e247f34ce924cd7304.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### OCR+IMAGES

CER: 8.67. Result ID: 69e736e247f34ce924cd7303.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### OCR+PAGEN

CER: 8.87. Result ID: 69e736e247f34ce924cd72ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### OCR+IMAGES

CER: 9.35. Result ID: 69e736e247f34ce924cd72fe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### OCR

CER: 9.54. Result ID: 69e736e247f34ce924cd7302.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### IMAGES

CER: 9.74. Result ID: 69e736e247f34ce924cd7301.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### IMAGES

CER: 9.87. Result ID: 69e736e247f34ce924cd72fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### IMAGES-ALL-AT-ONCE

CER: 9.88. Result ID: 69e736e247f34ce924cd7300.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMINI"}. This record does not assert independent reproduction.

### OCR

CER: 9.97. Result ID: 69e736e247f34ce924cd72fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### OCR+IMAGES

CER: 9.98. Result ID: 69e736e247f34ce924cd72f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

### IMAGES-ALL-AT-ONCE

CER: 10.18. Result ID: 69e736e247f34ce924cd72fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### OCR

CER: 10.75. Result ID: 69e736e247f34ce924cd72f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

### OCR+PAGE1

CER: 10.89. Result ID: 69e736e247f34ce924cd72f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

### OCR+PAGE1

CER: 10.95. Result ID: 69e736e247f34ce924cd72fa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4O"}. This record does not assert independent reproduction.

### OCR+PAGEN

CER: 11.02. Result ID: 69e736e247f34ce924cd72f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

### OCR

CER: 11.18. Result ID: 69e736e247f34ce924cd72f3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"Baseline"}. This record does not assert independent reproduction.

### IMAGES

CER: 15.6. Result ID: 69e736e247f34ce924cd72f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

### IMAGES-ALL-AT-ONCE

CER: 25.06. Result ID: 69e736e247f34ce924cd72f4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa772f1e209131cea71aa0.md>)

Recorded evaluation configuration: {"Base Model":"GEMMA"}. This record does not assert independent reproduction.

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