# Visual Question Answering on DocHaystack-1000

Source: [Visual Question Answering on DocHaystack-1000](<https://www.sota2.com/research/sota/visual-question-answering-on-dochaystack-1000>)

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

Data updated: 2026-02-26T05:30:13.917Z

Evaluation of model performance on the DocHaystack-1000 dataset for Visual Question Answering, typically measured using 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 | Visual Question Answering |
| Dataset (including any stated split) | DocHaystack-1000 |
| Metric | Accuracy |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/visual-question-answering-on-dochaystack-1000.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen2-VL-f.t. + V-RAG | 73.39 | Not recorded | 68fa76121e209131cea6b3d3 |
| GPT-4o + V-RAG | 66.97 | Not recorded | 68fa76121e209131cea6b3d3 |
| Qwen2-VL + V-RAG | 66.06 | Not recorded | 68fa76121e209131cea6b3d3 |
| Gemini + V-RAG | 58.72 | Not recorded | 68fa76121e209131cea6b3d3 |
| LLaVA-OV + V-RAG | 55.05 | Not recorded | 68fa76121e209131cea6b3d3 |
| MIRAGE | 2.75 | Not recorded | 68fa76121e209131cea6b3d3 |

### Qwen2-VL-f.t. + V-RAG

Accuracy: 73.39. Result ID: 699424263a6c3148491b581b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Implementation](<https://github.com/Vision-CAIR/dochaystacks>)

Recorded evaluation configuration: {"Framework":"V-RAG","Resolution":"Low-resolution","Adaptation":"LoRA"}. This record does not assert independent reproduction.

### GPT-4o + V-RAG

Accuracy: 66.97. Result ID: 699424263a6c3148491b5818.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Implementation](<https://github.com/Vision-CAIR/dochaystacks>)

Recorded evaluation configuration: {"Framework":"V-RAG","Resolution":"Low-resolution"}. This record does not assert independent reproduction.

### Qwen2-VL + V-RAG

Accuracy: 66.06. Result ID: 699424263a6c3148491b581a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Implementation](<https://github.com/Vision-CAIR/dochaystacks>)

Recorded evaluation configuration: {"Framework":"V-RAG","Resolution":"Low-resolution"}. This record does not assert independent reproduction.

### Gemini + V-RAG

Accuracy: 58.72. Result ID: 699424263a6c3148491b5819.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Implementation](<https://github.com/Vision-CAIR/dochaystacks>)

Recorded evaluation configuration: {"Framework":"V-RAG"}. This record does not assert independent reproduction.

### LLaVA-OV + V-RAG

Accuracy: 55.05. Result ID: 699424263a6c3148491b5817.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Implementation](<https://github.com/Vision-CAIR/dochaystacks>)

Recorded evaluation configuration: {"Framework":"V-RAG"}. This record does not assert independent reproduction.

### MIRAGE

Accuracy: 2.75. Result ID: 699424263a6c3148491b5816.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76121e209131cea6b3d3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77581e209131cea728c4.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.