# Spatial Reasoning on SPBench

Source: [Spatial Reasoning on SPBench](<https://www.sota2.com/research/sota/spatial-reasoning-on-spbench>)

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

Data updated: 2026-04-07T07:45:53.875Z

Evaluation of model performance on the SPBench dataset for the Spatial Reasoning task.

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 Reasoning |
| Dataset (including any stated split) | SPBench |
| Metric | SPBench Score |
| Unit | SCALAR |
| 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

- [SPBench Score](<https://www.sota2.com/research/sota/spatial-reasoning-on-spbench.md?metric=SPBench%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | SPBench Score | Result date | Report paper |
| --- | --- | --- | --- |
| Gemini 2.5 Pro | 55.9 | Not recorded | 69d48f997d0172140ec7434d |
| EgoMind | 55.02 | Not recorded | 69d48f997d0172140ec7434d |
| GPT-5 | 54.77 | Not recorded | 69d48f997d0172140ec7434d |
| GPT-4.1 | 54.51 | Not recorded | 69d48f997d0172140ec7434d |
| SpaceR | 53.39 | Not recorded | 69d48f997d0172140ec7434d |
| ViLaSR | 53.16 | Not recorded | 69d48f997d0172140ec7434d |
| InternVL3.5 | 52.23 | Not recorded | 69d48f997d0172140ec7434d |
| Spatial-MLLM | 48.4 | Not recorded | 69d48f997d0172140ec7434d |
| Qwen2.5-VL | 41.65 | Not recorded | 69d48f997d0172140ec7434d |
| MiniCPM-V-4.5 | 38.61 | Not recorded | 69d48f997d0172140ec7434d |
| MiMo-VL-RL | 33.3 | Not recorded | 69d48f997d0172140ec7434d |
| LLaVA-OneVision | 32.7 | Not recorded | 69d48f997d0172140ec7434d |
| Ovis2.5 | 25.75 | Not recorded | 69d48f997d0172140ec7434d |
| GPT4Scene | 23.52 | Not recorded | 69d48f997d0172140ec7434d |

### Gemini 2.5 Pro

SPBench Score: 55.9. Result ID: 69d4b5c647f34ce924bc3b4c.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

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

### EgoMind

SPBench Score: 55.02. Result ID: 69d4b5c647f34ce924bc3b57.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

Recorded evaluation configuration: {"Params":"7B","Data Size":"25k"}. This record does not assert independent reproduction.

### GPT-5

SPBench Score: 54.77. Result ID: 69d4b5c647f34ce924bc3b4b.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

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

### GPT-4.1

SPBench Score: 54.51. Result ID: 69d4b5c647f34ce924bc3b4a.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

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

### SpaceR

SPBench Score: 53.39. Result ID: 69d4b5c647f34ce924bc3b55.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a41e209131cea798a1.md>)

Recorded evaluation configuration: {"Params":"7B","Data Size":"151k"}. This record does not assert independent reproduction.

### ViLaSR

SPBench Score: 53.16. Result ID: 69d4b5c647f34ce924bc3b56.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79831e209131cea7defc.md>)

Recorded evaluation configuration: {"Params":"7B","Data Size":"80k"}. This record does not assert independent reproduction.

### InternVL3.5

SPBench Score: 52.23. Result ID: 69d4b5c647f34ce924bc3b4e.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

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

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

### Spatial-MLLM

SPBench Score: 48.4. Result ID: 69d4b5c647f34ce924bc3b54.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78f11e209131cea7b58a.md>)

Recorded evaluation configuration: {"Params":"4B","Data Size":"120K"}. This record does not assert independent reproduction.

### Qwen2.5-VL

SPBench Score: 41.65. Result ID: 69d4b5c647f34ce924bc3b4d.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

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

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

### MiniCPM-V-4.5

SPBench Score: 38.61. Result ID: 69d4b5c647f34ce924bc3b51.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b8d1e209131cea8913f.md>)

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

### MiMo-VL-RL

SPBench Score: 33.3. Result ID: 69d4b5c647f34ce924bc3b4f.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79121e209131cea7bfbd.md>)

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

### LLaVA-OneVision

SPBench Score: 32.7. Result ID: 69d4b5c647f34ce924bc3b52.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa717d1e209131cea51970.md>)

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

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

### Ovis2.5

SPBench Score: 25.75. Result ID: 69d4b5c647f34ce924bc3b50.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7aa61e209131cea84620.md>)

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

### GPT4Scene

SPBench Score: 23.52. Result ID: 69d4b5c647f34ce924bc3b53.

[Reporting paper](<https://www.sota2.com/research/paper/69d48f997d0172140ec7434d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77591e209131cea72932.md>)

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

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