# Mathematical Reasoning on AIME 2025 (Average@12 accuracy)

Source: [Mathematical Reasoning on AIME 2025 (Average@12 accuracy)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-2025-average-at-12-accuracy>)

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

Data updated: 2026-05-13T08:24:59.889Z

Evaluation of model performance on the American Invitational Mathematics Examination (AIME) 2025 dataset, reporting Average@12 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 | Mathematical Reasoning |
| Dataset (including any stated split) | AIME 2025 |
| Metric | Average@12 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

- [Average@12 Accuracy](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-2025-average-at-12-accuracy.md?metric=Average%4012%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average@12 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| ATESD | 72.5 | Not recorded | 6a0405badbc836b87d8276de |
| ATESD | 71.39 | Not recorded | 6a0405badbc836b87d8276de |
| OPSD | 70.8 | Not recorded | 6a0405badbc836b87d8276de |
| GRPO | 68.9 | Not recorded | 6a0405badbc836b87d8276de |
| OPSD | 68.3 | Not recorded | 6a0405badbc836b87d8276de |
| GRPO | 68.1 | Not recorded | 6a0405badbc836b87d8276de |
| Base (Instruct) | 66.4 | Not recorded | 6a0405badbc836b87d8276de |
| Base (Instruct) | 65.6 | Not recorded | 6a0405badbc836b87d8276de |
| SFT | 64.2 | Not recorded | 6a0405badbc836b87d8276de |
| SFT | 62.3 | Not recorded | 6a0405badbc836b87d8276de |
| ATESD | 44.72 | Not recorded | 6a0405badbc836b87d8276de |
| OPSD | 43.9 | Not recorded | 6a0405badbc836b87d8276de |
| GRPO | 38.3 | Not recorded | 6a0405badbc836b87d8276de |
| Base (Instruct) | 36.7 | Not recorded | 6a0405badbc836b87d8276de |
| SFT | 36.3 | Not recorded | 6a0405badbc836b87d8276de |

### ATESD

Average@12 Accuracy: 72.5. Result ID: 6a04354e47f34ce924ea110f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### ATESD

Average@12 Accuracy: 71.39. Result ID: 6a04354e47f34ce924ea110a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### OPSD

Average@12 Accuracy: 70.8. Result ID: 6a04354e47f34ce924ea110e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6978471ca2b28e761d87ca01.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### GRPO

Average@12 Accuracy: 68.9. Result ID: 6a04354e47f34ce924ea110d.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### OPSD

Average@12 Accuracy: 68.3. Result ID: 6a04354e47f34ce924ea1109.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6978471ca2b28e761d87ca01.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### GRPO

Average@12 Accuracy: 68.1. Result ID: 6a04354e47f34ce924ea1108.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Base (Instruct)

Average@12 Accuracy: 66.4. Result ID: 6a04354e47f34ce924ea1106.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Base (Instruct)

Average@12 Accuracy: 65.6. Result ID: 6a04354e47f34ce924ea110b.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### SFT

Average@12 Accuracy: 64.2. Result ID: 6a04354e47f34ce924ea110c.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### SFT

Average@12 Accuracy: 62.3. Result ID: 6a04354e47f34ce924ea1107.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### ATESD

Average@12 Accuracy: 44.72. Result ID: 6a04354e47f34ce924ea1105.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### OPSD

Average@12 Accuracy: 43.9. Result ID: 6a04354e47f34ce924ea1104.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6978471ca2b28e761d87ca01.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GRPO

Average@12 Accuracy: 38.3. Result ID: 6a04354e47f34ce924ea1103.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Base (Instruct)

Average@12 Accuracy: 36.7. Result ID: 6a04354e47f34ce924ea1101.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### SFT

Average@12 Accuracy: 36.3. Result ID: 6a04354e47f34ce924ea1102.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405badbc836b87d8276de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

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