# Math on AIME 2025 (Acc, AMSR)

Source: [Math on AIME 2025 (Acc, AMSR)](<https://www.sota2.com/research/sota/math-on-aime-2025-acc-amsr>)

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

Data updated: 2026-05-15T07:04:10.506Z

Evaluation of model performance on the American Invitational Mathematics Examination (AIME) 2025 dataset, reporting Accuracy (Acc) and AMSR metrics.

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 | Math |
| Dataset (including any stated split) | AIME 2025 |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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/math-on-aime-2025-acc-amsr.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [AMSR](<https://www.sota2.com/research/sota/math-on-aime-2025-acc-amsr.md?metric=AMSR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Dense | 67.7 | Not recorded | 68fa7a521e209131cea826ed |
| ReasonCache | 63.3 | Not recorded | 68fa7a521e209131cea826ed |
| Dense | 63.3 | Not recorded | 68fa7a521e209131cea826ed |
| ReasonCache | 63.3 | Not recorded | 68fa7a521e209131cea826ed |
| Dense | 53.3 | Not recorded | 68fa7a521e209131cea826ed |
| Quest | 53.3 | Not recorded | 68fa7a521e209131cea826ed |
| ReasonCache | 50 | Not recorded | 68fa7a521e209131cea826ed |
| Quest | 43.3 | Not recorded | 68fa7a521e209131cea826ed |
| StreamingLLM | 36.7 | Not recorded | 68fa7a521e209131cea826ed |
| StreamingLLM | 36.7 | Not recorded | 68fa7a521e209131cea826ed |
| StreamingLLM | 33.3 | Not recorded | 68fa7a521e209131cea826ed |
| Quest | 33.3 | Not recorded | 68fa7a521e209131cea826ed |
| SnapKV | 23.3 | Not recorded | 68fa7a521e209131cea826ed |
| SnapKV | 20 | Not recorded | 68fa7a521e209131cea826ed |
| SnapKV | 16.6 | Not recorded | 68fa7a521e209131cea826ed |

### Dense

Accuracy: 67.7. Result ID: 6a06c56647f34ce924edd8d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"Phi 4 reasoning plus"}. This record does not assert independent reproduction.

### ReasonCache

Accuracy: 63.3. Result ID: 6a06c56647f34ce924edd8cc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"QwQ 32B"}. This record does not assert independent reproduction.

### Dense

Accuracy: 63.3. Result ID: 6a06c56647f34ce924edd8cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"QwQ 32B"}. This record does not assert independent reproduction.

### ReasonCache

Accuracy: 63.3. Result ID: 6a06c56647f34ce924edd8d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"Phi 4 reasoning plus"}. This record does not assert independent reproduction.

### Dense

Accuracy: 53.3. Result ID: 6a06c56647f34ce924edd8c8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"DeepSeek R1 Distill Qwen 32B"}. This record does not assert independent reproduction.

### Quest

Accuracy: 53.3. Result ID: 6a06c56647f34ce924edd8cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e6e.md>)

Recorded evaluation configuration: {"Model":"QwQ 32B"}. This record does not assert independent reproduction.

### ReasonCache

Accuracy: 50. Result ID: 6a06c56647f34ce924edd8c7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

Recorded evaluation configuration: {"Model":"DeepSeek R1 Distill Qwen 32B"}. This record does not assert independent reproduction.

### Quest

Accuracy: 43.3. Result ID: 6a06c56647f34ce924edd8d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e6e.md>)

Recorded evaluation configuration: {"Model":"Phi 4 reasoning plus"}. This record does not assert independent reproduction.

### StreamingLLM

Accuracy: 36.7. Result ID: 6a06c56647f34ce924edd8c9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Model":"QwQ 32B"}. This record does not assert independent reproduction.

### StreamingLLM

Accuracy: 36.7. Result ID: 6a06c56647f34ce924edd8ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Model":"Phi 4 reasoning plus"}. This record does not assert independent reproduction.

### StreamingLLM

Accuracy: 33.3. Result ID: 6a06c56647f34ce924edd8c4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Model":"DeepSeek R1 Distill Qwen 32B"}. This record does not assert independent reproduction.

### Quest

Accuracy: 33.3. Result ID: 6a06c56647f34ce924edd8c6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e6e.md>)

Recorded evaluation configuration: {"Model":"DeepSeek R1 Distill Qwen 32B"}. This record does not assert independent reproduction.

### SnapKV

Accuracy: 23.3. Result ID: 6a06c56647f34ce924edd8ca.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Model":"QwQ 32B"}. This record does not assert independent reproduction.

### SnapKV

Accuracy: 20. Result ID: 6a06c56647f34ce924edd8c5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Model":"DeepSeek R1 Distill Qwen 32B"}. This record does not assert independent reproduction.

### SnapKV

Accuracy: 16.6. Result ID: 6a06c56647f34ce924edd8cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a521e209131cea826ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Model":"Phi 4 reasoning plus"}. This record does not assert independent reproduction.

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