# Spatial VQA on 3DSRBench Level 1-3

Source: [Spatial VQA on 3DSRBench Level 1-3](<https://www.sota2.com/research/sota/spatial-vqa-on-3dsrbench-level-1-3>)

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

Data updated: 2026-03-27T13:21:44.192Z

Evaluation of model performance on Levels 1 through 3 of the 3DSRBench dataset, focusing on Spatial Visual Question Answering capabilities, 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 | Spatial VQA |
| Dataset (including any stated split) | 3DSRBench Level 1-3 |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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/spatial-vqa-on-3dsrbench-level-1-3.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| HiSpatial-3B-RGB | 64.34 | Not recorded | 69c60f310636e5bf4afc865b |
| HiSpatial-3B | 63.81 | Not recorded | 69c60f310636e5bf4afc865b |
| InternVL3.5-14B | 57.32 | Not recorded | 69c60f310636e5bf4afc865b |
| Qwen-3-VL-30B-A3B | 55.7 | Not recorded | 69c60f310636e5bf4afc865b |
| SpatialBot-3B | 54.67 | Not recorded | 69c60f310636e5bf4afc865b |
| Qwen-3-VL-8B | 52.8 | Not recorded | 69c60f310636e5bf4afc865b |
| InternVL3.5-8B | 51.58 | Not recorded | 69c60f310636e5bf4afc865b |
| RoboRefer-8B-SFT | 51.24 | Not recorded | 69c60f310636e5bf4afc865b |
| Gemini-2.5-Pro | 48.47 | Not recorded | 69c60f310636e5bf4afc865b |
| SpatialRGPT-8B | 48.4 | Not recorded | 69c60f310636e5bf4afc865b |
| Claude-3.7-Sonnet | 48.2 | Not recorded | 69c60f310636e5bf4afc865b |
| SpaceLLaVA-13B | 46.55 | Not recorded | 69c60f310636e5bf4afc865b |
| GPT-4o | 44.2 | Not recorded | 69c60f310636e5bf4afc865b |
| PaliGemma2-3B | 8.15 | Not recorded | 69c60f310636e5bf4afc865b |

### HiSpatial-3B-RGB

Accuracy: 64.34. Result ID: 69c6845447f34ce924b0a0eb.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### HiSpatial-3B

Accuracy: 63.81. Result ID: 69c6845447f34ce924b0a0ec.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

Recorded evaluation configuration: {"Input":"RGB-XYZ"}. This record does not assert independent reproduction.

### InternVL3.5-14B

Accuracy: 57.32. Result ID: 69c6845447f34ce924b0a0e4.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### Qwen-3-VL-30B-A3B

Accuracy: 55.7. Result ID: 69c6845447f34ce924b0a0e6.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### SpatialBot-3B

Accuracy: 54.67. Result ID: 69c6845447f34ce924b0a0e7.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77821e209131cea73b63.md>)

[Implementation](<https://github.com/BAAI-DCAI/SpatialBot>)

Recorded evaluation configuration: {"Input":"RGB-D"}. This record does not assert independent reproduction.

### Qwen-3-VL-8B

Accuracy: 52.8. Result ID: 69c6845447f34ce924b0a0e5.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### InternVL3.5-8B

Accuracy: 51.58. Result ID: 69c6845447f34ce924b0a0e3.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### RoboRefer-8B-SFT

Accuracy: 51.24. Result ID: 69c6845447f34ce924b0a0ea.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c401e209131cea8d2b8.md>)

Recorded evaluation configuration: {"Input":"RGB-D"}. This record does not assert independent reproduction.

### Gemini-2.5-Pro

Accuracy: 48.47. Result ID: 69c6845447f34ce924b0a0e0.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c311e209131cea8cef5.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### SpatialRGPT-8B

Accuracy: 48.4. Result ID: 69c6845447f34ce924b0a0e9.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75321e209131cea66538.md>)

[Implementation](<https://www.anjiecheng.me/SpatialRGPT>)

Recorded evaluation configuration: {"Input":"RGB-D"}. This record does not assert independent reproduction.

### Claude-3.7-Sonnet

Accuracy: 48.2. Result ID: 69c6845447f34ce924b0a0e1.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### SpaceLLaVA-13B

Accuracy: 46.55. Result ID: 69c6845447f34ce924b0a0e8.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71f91e209131cea53d86.md>)

[Implementation](<https://spatial-vlm.github.io/>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### GPT-4o

Accuracy: 44.2. Result ID: 69c6845447f34ce924b0a0df.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

### PaliGemma2-3B

Accuracy: 8.15. Result ID: 69c6845447f34ce924b0a0e2.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f310636e5bf4afc865b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76001e209131cea6ad66.md>)

Recorded evaluation configuration: {"Input":"RGB"}. This record does not assert independent reproduction.

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