# Open-ended Question Answering on CountBench

Source: [Open-ended Question Answering on CountBench](<https://www.sota2.com/research/sota/open-ended-question-answering-on-countbench>)

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

Data updated: 2026-06-04T10:04:07.340Z

Evaluation of model performance on the Open-ended Question Answering task using the CountBench dataset, measured by Accuracy.

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 | Open-ended Question Answering |
| Dataset (including any stated split) | CountBench |
| Metric | Accuracy (%) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [Accuracy (%)](<https://www.sota2.com/research/sota/open-ended-question-answering-on-countbench.md?metric=Accuracy%20(%25)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (%) | Result date | Report paper |
| --- | --- | --- | --- |
| OneVision-8B | 85.5 | Not recorded | 6a210692c10484bac23f01c1 |
| OneVision-4B | 83.7 | Not recorded | 6a210692c10484bac23f01c1 |
| Llama-90B | 78 | Not recorded | 6a210692c10484bac23f01c1 |
| OneVision-4B-LN | 74.9 | Not recorded | 6a210692c10484bac23f01c1 |
| InternVL2-8B | 58.5 | Not recorded | 6a210692c10484bac23f01c1 |
| LLaVA-13B | 48.3 | Not recorded | 6a210692c10484bac23f01c1 |
| LLaVA-7B | 45.8 | Not recorded | 6a210692c10484bac23f01c1 |
| Cambrian-8B | 15.1 | Not recorded | 6a210692c10484bac23f01c1 |

### OneVision-8B

Accuracy (%): 85.5. Result ID: 6a214d88fbc373592c1d7b8b.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

Recorded evaluation configuration: {"Model":"OneVision-8B"}. This record does not assert independent reproduction.

### OneVision-4B

Accuracy (%): 83.7. Result ID: 6a214d88fbc373592c1d7b8a.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

Recorded evaluation configuration: {"Model":"OneVision-4B"}. This record does not assert independent reproduction.

### Llama-90B

Accuracy (%): 78. Result ID: 6a214d88fbc373592c1d7b8e.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

Recorded evaluation configuration: {"Model":"Llama-90B"}. This record does not assert independent reproduction.

### OneVision-4B-LN

Accuracy (%): 74.9. Result ID: 6a214d88fbc373592c1d7b89.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

Recorded evaluation configuration: {"Model":"OneVision-4B-LN"}. This record does not assert independent reproduction.

### InternVL2-8B

Accuracy (%): 58.5. Result ID: 6a214d88fbc373592c1d7b8c.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

Recorded evaluation configuration: {"Model":"InternVL2-8B"}. This record does not assert independent reproduction.

### LLaVA-13B

Accuracy (%): 48.3. Result ID: 6a214d88fbc373592c1d7b88.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa732e1e209131cea5af59.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Model":"LLaVA-13B"}. This record does not assert independent reproduction.

### LLaVA-7B

Accuracy (%): 45.8. Result ID: 6a214d88fbc373592c1d7b87.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa732e1e209131cea5af59.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Model":"LLaVA-7B"}. This record does not assert independent reproduction.

### Cambrian-8B

Accuracy (%): 15.1. Result ID: 6a214d88fbc373592c1d7b8d.

[Reporting paper](<https://www.sota2.com/research/paper/6a210692c10484bac23f01c1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac2e.md>)

Recorded evaluation configuration: {"Model":"Cambrian-8B"}. This record does not assert independent reproduction.

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