# AIME 2025

> Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/aime-2025. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/aime-2025.

AIME 2025: Solving final-answer mathematics competition problems. This competition contains 30 problems. MathArena reports normalized performance across repeated attempts and a 95% confidence interval.

Canonical URL: https://benchize.com/benchmarks/aime-2025
Last updated: Wed, 21 Oct 2026 18:52:20 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

Conquered

AIME 2025: Solving final-answer mathematics competition problems. This competition contains 30 problems. MathArena reports normalized performance across repeated attempts and a 95% confidence interval.

Mathematics · Models tested: 49 · Score publishers: [matharena](https://matharena.ai/?comp=aime--aime_2025), Vals AI

## Model results

| # | Model | Score | ± | Cost | Latency | Source |  |
| --- | --- | --- | --- | --- | --- | --- | --- |
| 1 / 35 | [GPT-5.2](https://benchize.com/models/gpt-5-2-high.md)High | 100% | 95% CI 100–100% | 0.11usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 100.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.11 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT-5.2 (high) sourceModelVersion openai/gpt-52-high thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 100–100% confidenceInterval95 0.00 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5.2 (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/0` |  |  |  |  |  |  |  |
| 1 / 35 | [GPT-5.2](https://benchize.com/models/gpt-5-2-xhigh.md)Extra high | 100% | 95% CI 100–100% | 0.18usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 100.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.18 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort xhigh sourceModelName GPT-5.2 (xhigh) sourceModelVersion openai/gpt-52 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 100–100% confidenceInterval95 0.00 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5.2 (xhigh) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/1` |  |  |  |  |  |  |  |
| 3 / 35 | [Step 3.5 Flash](https://benchize.com/models/step-3-5-flash.md) | 98.33% | 95% CI 96.04–100.62% | 0.01usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 98.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.011 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Step 3.5 Flash sourceModelVersion stepfun/3.5-flash ### Reported statistics Uncertainty 95% confidence interval: 96.04–100.62% confidenceInterval95 2.29 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Step 3.5 Flash Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/2` |  |  |  |  |  |  |  |
| 4 / 35 | [Gemini 3 Flash](https://benchize.com/models/gemini-3-flash.md) | 97.5% | 95% CI 94.71–100.29% | 0.06usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 97.50 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.055 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Gemini 3 Flash sourceModelVersion gemini/gemini-3-flash ### Reported statistics Uncertainty 95% confidence interval: 94.71–100.29% confidenceInterval95 2.79 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Gemini 3 Flash Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/3` |  |  |  |  |  |  |  |
| 1 / 14 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview-high.md)High | 97.08% | SE ±0.42 pp | $0.14 | 115.73source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 97.083 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.142586 \| USD \| per_test \| Lower is better \| \| `latency` \| 115.726 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.42 pp sourcePopulation 96 stderr 0.417 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label google/gemini-3.1-pro-preview Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/google~1gemini-3.1-pro-preview` |  |  |  |  |  |  |  |
| 5 / 35 | [GLM-5](https://benchize.com/models/glm-5.md) | 96.67% | 95% CI 93.46–99.88% | 0.08usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 96.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.081 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName GLM 5 sourceModelVersion glm/glm-5 ### Reported statistics Uncertainty 95% confidence interval: 93.46–99.88% confidenceInterval95 3.21 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GLM 5 Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/4` |  |  |  |  |  |  |  |
| 2 / 14 | [Claude Opus 4.7](https://benchize.com/models/claude-opus-4-7.md) | 95.83% | SE ±1.05 pp | $0.56 | 227.42source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.833 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.561248 \| USD \| per_test \| Lower is better \| \| `latency` \| 227.418 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.05 pp sourcePopulation 96 stderr 1.045 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label anthropic/claude-opus-4-7 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/anthropic~1claude-opus-4-7` |  |  |  |  |  |  |  |
| 3 / 14 | [Gemini 3 Flash Preview](https://benchize.com/models/gemini-3-flash-preview-high.md)High | 95.42% | SE ±0.61 pp | $0.05 | 87.22source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.417 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.048415 \| USD \| per_test \| Lower is better \| \| `latency` \| 87.215 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.61 pp sourcePopulation 96 stderr 0.61 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label google/gemini-3-flash-preview Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/google~1gemini-3-flash-preview` |  |  |  |  |  |  |  |
| 6 / 35 | [GPT-5](https://benchize.com/models/gpt-5-high.md)High | 95% | 95% CI 91.1–98.9% | 0.14usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.14 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT-5 (high) sourceModelVersion openai/gpt-5 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 91.1–98.9% confidenceInterval95 3.90 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5 (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/7` |  |  |  |  |  |  |  |
| 7 / 35 | [GPT-5.1](https://benchize.com/models/gpt-5-1-high.md)High | 94.17% | 95% CI 89.98–98.36% | 0.18usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 94.17 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.18 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT-5.1 (high) sourceModelVersion openai/gpt-51 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 89.98–98.36% confidenceInterval95 4.19 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5.1 (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/10` |  |  |  |  |  |  |  |
| 8 / 35 | [GLM-4.5](https://benchize.com/models/glm-4-5.md) | 93.33% | 95% CI 88.87–97.79% | 0.05usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 93.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.048 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName GLM 4.5 sourceModelVersion glm/glm-45 ### Reported statistics Uncertainty 95% confidence interval: 88.87–97.79% confidenceInterval95 4.46 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GLM 4.5 Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/11` |  |  |  |  |  |  |  |
| 9 / 35 | [GLM-4.6](https://benchize.com/models/glm-4-6.md) | 91.67% | 95% CI 86.72–96.62% | 0.04usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.036 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName GLM 4.6 sourceModelVersion glm/glm-46 ### Reported statistics Uncertainty 95% confidence interval: 86.72–96.62% confidenceInterval95 4.95 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GLM 4.6 Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/15` |  |  |  |  |  |  |  |
| 4 / 14 | [Claude Sonnet 4.6](https://benchize.com/models/claude-sonnet-4-6.md) | 91.25% | SE ±0.88 pp | $0.52 | 388.99source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.25 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.520585 \| USD \| per_test \| Lower is better \| \| `latency` \| 388.987 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Anthropic reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.88 pp sourcePopulation 96 stderr 0.877 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label anthropic/claude-sonnet-4-6 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/anthropic~1claude-sonnet-4-6` |  |  |  |  |  |  |  |
| 5 / 14 | [MiniMax M2.7](https://benchize.com/models/minimax-m2-7.md) | 90.42% | SE ±0.76 pp | $0.02 | 236.76source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.417 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.023157 \| USD \| per_test \| Lower is better \| \| `latency` \| 236.759 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.76 pp sourcePopulation 96 stderr 0.755 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label minimax/MiniMax-M2.7 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/minimax~1MiniMax-M2.7` |  |  |  |  |  |  |  |
| 10 / 35 | [gpt-oss-120b](https://benchize.com/models/gpt-oss-120b-high.md)High | 90% | 95% CI 84.63–95.37% | 0.01usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.013 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT OSS 120B (high) sourceModelVersion openai/oss-120b thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 84.63–95.37% confidenceInterval95 5.37 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT OSS 120B (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/19` |  |  |  |  |  |  |  |
| 11 / 35 | [R1 0528](https://benchize.com/models/deepseek-r1-0528.md) | 89.17% | 95% CI 83.61–94.73% | 0.05usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.17 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.048 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName DeepSeek-R1-0528 sourceModelVersion deepseek/deepseek_r1_0528 ### Reported statistics Uncertainty 95% confidence interval: 83.61–94.73% confidenceInterval95 5.56 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label DeepSeek-R1-0528 Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/22` |  |  |  |  |  |  |  |
| 11 / 35 | [O3](https://benchize.com/models/o3-high.md)High | 89.17% | 95% CI 83.61–94.73% | 0.1usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.17 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.10 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName o3 (high) sourceModelVersion openai/o3 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 83.61–94.73% confidenceInterval95 5.56 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o3 (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/23` |  |  |  |  |  |  |  |
| 11 / 35 | [gpt-oss-20b](https://benchize.com/models/gpt-oss-20b-high.md)High | 89.17% | 95% CI 83.61–94.73% | 0.008usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.17 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.008 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT OSS 20B (high) sourceModelVersion openai/oss-20b thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 83.61–94.73% confidenceInterval95 5.56 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT OSS 20B (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/21` |  |  |  |  |  |  |  |
| 6 / 14 | [MiniMax M2.5](https://benchize.com/models/minimax-m2-5.md) | 88.33% | SE ±1.09 pp | $0.04 | 403.29source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.333 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.044437 \| USD \| per_test \| Lower is better \| \| `latency` \| 403.292 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.09 pp sourcePopulation 96 stderr 1.091 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label minimax/MiniMax-M2.5 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/minimax~1MiniMax-M2.5` |  |  |  |  |  |  |  |
| 14 / 35 | [Gemini 2.5 Pro](https://benchize.com/models/gemini-2-5-pro.md) | 88.33% | 95% CI 82.59–94.07% | 0.13usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.13 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Gemini 2.5 Pro sourceModelVersion gemini/gemini-pro-2.5 ### Reported statistics Uncertainty 95% confidence interval: 82.59–94.07% confidenceInterval95 5.74 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Gemini 2.5 Pro Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/24` |  |  |  |  |  |  |  |
| 15 / 35 | [GPT-5 Mini](https://benchize.com/models/gpt-5-mini-high.md)High | 87.5% | 95% CI 81.58–93.42% | 0.03usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.50 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.033 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT-5-mini (high) sourceModelVersion openai/gpt-5-mini thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 81.58–93.42% confidenceInterval95 5.92 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5-mini (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/25` |  |  |  |  |  |  |  |
| 16 / 35 | [O3 Mini](https://benchize.com/models/o3-mini-high.md)High | 86.67% | 95% CI 80.59–92.75% | 0.05usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.050 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName o3-mini (high) sourceModelVersion openai/o3-mini thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 80.59–92.75% confidenceInterval95 6.08 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o3-mini (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/27` |  |  |  |  |  |  |  |
| 17 / 35 | [GPT-5 Nano](https://benchize.com/models/gpt-5-nano-high.md)High | 85% | 95% CI 78.61–91.39% | 0.01usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 85.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.011 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName GPT-5-nano (high) sourceModelVersion openai/gpt-5-nano thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 78.61–91.39% confidenceInterval95 6.39 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5-nano (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/28` |  |  |  |  |  |  |  |
| 18 / 35 | [O4 Mini](https://benchize.com/models/o4-mini-medium.md)Medium | 84.17% | 95% CI 77.64–90.7% | 0.03usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 84.17 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.028 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort medium sourceModelName o4-mini (medium) sourceModelVersion openai/o4-mini--medium thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 77.64–90.7% confidenceInterval95 6.53 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o4-mini (medium) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/29` |  |  |  |  |  |  |  |
| 19 / 35 | [GLM-4.5 Air](https://benchize.com/models/glm-4-5-air.md) | 83.33% | 95% CI 76.66–90% | 0.03usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 83.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.026 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName GLM 4.5 Air sourceModelVersion glm/glm-45-air ### Reported statistics Uncertainty 95% confidence interval: 76.66–90% confidenceInterval95 6.67 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GLM 4.5 Air Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/33` |  |  |  |  |  |  |  |
| 20 / 35 | [O1](https://benchize.com/models/o1-medium.md)Medium | 81.67% | 95% CI 74.75–88.59% | 0.71usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.71 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort medium sourceModelName o1 (medium) sourceModelVersion openai/o1 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 74.75–88.59% confidenceInterval95 6.92 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o1 (medium) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/35` |  |  |  |  |  |  |  |
| 20 / 35 | [Grok 3 Mini](https://benchize.com/models/grok-3-mini-high.md)High | 81.67% | 95% CI 74.75–88.59% | 0.009usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.009 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort high sourceModelName Grok 3 Mini (high) sourceModelVersion xai/grok-3-mini-high thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 74.75–88.59% confidenceInterval95 6.92 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Grok 3 Mini (high) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/34` |  |  |  |  |  |  |  |
| 22 / 35 | [Qwen3 235B A22B](https://benchize.com/models/qwen3-235b-a22b.md) | 80.83% | 95% CI 73.79–87.87% | 0.009usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80.83 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.009 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Qwen3-235B-A22B sourceModelVersion qwen/qwen3_235b_a22b ### Reported statistics Uncertainty 95% confidence interval: 73.79–87.87% confidenceInterval95 7.04 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Qwen3-235B-A22B Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/36` |  |  |  |  |  |  |  |
| 7 / 14 | [Gemini 3.1 Flash Lite Preview](https://benchize.com/models/gemini-3-1-flash-lite-preview-high.md)High | 78.33% | SE ±2.18 pp | $0.00744 | 42.97source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 78.333 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.007442 \| USD \| per_test \| Lower is better \| \| `latency` \| 42.968 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.18 pp sourcePopulation 96 stderr 2.182 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label google/gemini-3.1-flash-lite-preview Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/google~1gemini-3.1-flash-lite-preview` |  |  |  |  |  |  |  |
| 23 / 35 | [O3 Mini](https://benchize.com/models/o3-mini-medium.md)Medium | 76.67% | 95% CI 69.1–84.24% | 0.03usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 76.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.027 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort medium sourceModelName o3-mini (medium) sourceModelVersion openai/o3-mini--medium thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 69.1–84.24% confidenceInterval95 7.57 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o3-mini (medium) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/38` |  |  |  |  |  |  |  |
| 8 / 14 | [MiniMax M2.1](https://benchize.com/models/minimax-m2-1.md) | 76.67% | SE ±1.89 pp | $0.03 | 257.33source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 76.667 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.025666 \| USD \| per_test \| Lower is better \| \| `latency` \| 257.333 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.89 pp sourcePopulation 96 stderr 1.89 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label minimax/MiniMax-M2.1 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/minimax~1MiniMax-M2.1` |  |  |  |  |  |  |  |
| 24 / 35 | [Gemini 2.5 Flash](https://benchize.com/models/gemini-2-5-flash-thinking.md)Thinking | 70.83% | 95% CI 62.7–78.96% | 0.08usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 70.83 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.084 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Gemini 2.5 Flash (Thinking) sourceModelVersion gemini/gemini-flash-2.5 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 62.7–78.96% confidenceInterval95 8.13 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Gemini 2.5 Flash (Thinking) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/39` |  |  |  |  |  |  |  |
| 25 / 35 | [Qwen3 30B A3B](https://benchize.com/models/qwen3-30b-a3b.md) | 70% | 95% CI 61.8–78.2% | 0.005usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 70.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.005 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Qwen3-30B-A3B sourceModelVersion qwen/qwen3_30b_a3b ### Reported statistics Uncertainty 95% confidence interval: 61.8–78.2% confidenceInterval95 8.20 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Qwen3-30B-A3B Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/40` |  |  |  |  |  |  |  |
| 25 / 35 | [Claude Opus 4](https://benchize.com/models/claude-opus-4.md) | 70% | 95% CI 61.8–78.2% | 1.13usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 70.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 1.13 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Claude-Opus-4.0 (Think) sourceModelVersion anthropic/claude-opus-4 ### Reported statistics Uncertainty 95% confidence interval: 61.8–78.2% confidenceInterval95 8.20 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Claude-Opus-4.0 (Think) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/42` |  |  |  |  |  |  |  |
| 25 / 35 | [DeepSeek R1](https://benchize.com/models/deepseek-r1.md) | 70% | 95% CI 61.8–78.2% | 0.03usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 70.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.025 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName DeepSeek-R1 sourceModelVersion deepseek/deepseek_r1 ### Reported statistics Uncertainty 95% confidence interval: 61.8–78.2% confidenceInterval95 8.20 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label DeepSeek-R1 Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/41` |  |  |  |  |  |  |  |
| 28 / 35 | [Grok 3 Mini](https://benchize.com/models/grok-3-mini-low.md)Low | 65% | 95% CI 56.47–73.53% | 0.003usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 65.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.003 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort low sourceModelName Grok 3 Mini (low) sourceModelVersion xai/grok-3-mini-low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 56.47–73.53% confidenceInterval95 8.53 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Grok 3 Mini (low) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/44` |  |  |  |  |  |  |  |
| 29 / 35 | [O4 Mini](https://benchize.com/models/o4-mini-low.md)Low | 61.67% | 95% CI 52.97–70.37% | 0.01usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 61.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.012 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort low sourceModelName o4-mini (low) sourceModelVersion openai/o4-mini--low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 52.97–70.37% confidenceInterval95 8.70 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o4-mini (low) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/45` |  |  |  |  |  |  |  |
| 30 / 35 | [O3 Mini](https://benchize.com/models/o3-mini-low.md)Low | 48.33% | 95% CI 39.39–57.27% | 0.01usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 48.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.011 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort low sourceModelName o3-mini (low) sourceModelVersion openai/o3-mini--low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 39.39–57.27% confidenceInterval95 8.94 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label o3-mini (low) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/52` |  |  |  |  |  |  |  |
| 31 / 35 | [GPT-5 Nano](https://benchize.com/models/gpt-5-nano-low.md)Low | 47.22% | 95% CI 38.17–56.27% | 0.001usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 47.22 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.001 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false Reasoning effort low sourceModelName GPT-5-nano (low) sourceModelVersion openai/gpt-5-nano-low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 38.17–56.27% confidenceInterval95 9.05 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label GPT-5-nano (low) Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/53` |  |  |  |  |  |  |  |
| 9 / 14 | [Mistral Large 3 2512](https://benchize.com/models/mistral-large-2512.md) | 31.25% | SE ±1.66 pp | $0.00631 | 106.8source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 31.25 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006307 \| USD \| per_test \| Lower is better \| \| `latency` \| 106.797 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Mistral AI reasoning Not reported Reasoning effort unknown Temperature 0.065 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.66 pp sourcePopulation 96 stderr 1.659 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label mistralai/mistral-large-2512 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/mistralai~1mistral-large-2512` |  |  |  |  |  |  |  |
| 32 / 35 | [Gemini 2.0 Pro](https://benchize.com/models/gemini-2-0-pro.md) | 27.5% | 95% CI 19.51–35.49% | 0.02usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 27.50 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.015 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName gemini-2.0-pro sourceModelVersion gemini/gemini-pro-2.0 ### Reported statistics Uncertainty 95% confidence interval: 19.51–35.49% confidenceInterval95 7.99 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label gemini-2.0-pro Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/56` |  |  |  |  |  |  |  |
| 32 / 35 | [Gemini 2.0 Flash](https://benchize.com/models/gemini-2-0-flash.md) | 27.5% | 95% CI 19.51–35.49% | 0.001usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 27.50 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.001 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName gemini-2.0-flash sourceModelVersion gemini/gemini-flash-2.0 ### Reported statistics Uncertainty 95% confidence interval: 19.51–35.49% confidenceInterval95 7.99 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label gemini-2.0-flash Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/55` |  |  |  |  |  |  |  |
| 10 / 14 | [Gemini 2.5 Flash Lite Preview 09-2025](https://benchize.com/models/gemini-2-5-flash-lite-preview-09-2025.md) | 21.25% | SE ±3.33 pp | $0.00352 | 20.31source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 21.25 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.003516 \| USD \| per_test \| Lower is better \| \| `latency` \| 20.313 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Google reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±3.33 pp sourcePopulation 96 stderr 3.33 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label google/gemini-2.5-flash-lite-preview-09-2025 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/google~1gemini-2.5-flash-lite-preview-09-2025` |  |  |  |  |  |  |  |
| 11 / 14 | [GPT-4o (2024-08-06)](https://benchize.com/models/gpt-4o-2024-08-06-high.md)High | 15% | SE ±1.26 pp | $0.01 | 35.23source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 15 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.010184 \| USD \| per_test \| Lower is better \| \| `latency` \| 35.226 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.26 pp sourcePopulation 96 stderr 1.26 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label openai/gpt-4o-2024-08-06 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/openai~1gpt-4o-2024-08-06` |  |  |  |  |  |  |  |
| 12 / 14 | [GPT-4o-mini (2024-07-18)](https://benchize.com/models/gpt-4o-mini-2024-07-18-high.md)High | 14.17% | SE ±1.37 pp | $0.000678 | 17.1source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 14.167 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000678 \| USD \| per_test \| Lower is better \| \| `latency` \| 17.102 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.37 pp sourcePopulation 96 stderr 1.373 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label openai/gpt-4o-mini-2024-07-18 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/openai~1gpt-4o-mini-2024-07-18` |  |  |  |  |  |  |  |
| 34 / 35 | [GPT-4o](https://benchize.com/models/gpt-4o.md) | 11.67% | 95% CI 5.93–17.41% | 0.009usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 11.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.009 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName gpt-4o sourceModelVersion openai/gpt-4o ### Reported statistics Uncertainty 95% confidence interval: 5.93–17.41% confidenceInterval95 5.74 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label gpt-4o Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/59` |  |  |  |  |  |  |  |
| 13 / 14 | [Command A](https://benchize.com/models/command-a.md) | 11.67% | SE ±1.09 pp | $0.01 | 23.2source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 11.667 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.010479 \| USD \| per_test \| Lower is better \| \| `latency` \| 23.196 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider Cohere reasoning Not reported Reasoning effort unknown Temperature 0.3 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.09 pp sourcePopulation 96 stderr 1.091 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label cohere/command-a-03-2025 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/cohere~1command-a-03-2025` |  |  |  |  |  |  |  |
| 14 / 14 | [GPT-4o (2024-11-20)](https://benchize.com/models/gpt-4o-2024-11-20-high.md)High | 9.58% | SE ±1.6 pp | $0.01 | 15.59source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 9.583 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.011992 \| USD \| per_test \| Lower is better \| \| `latency` \| 15.593 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens Not reported provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.6 pp sourcePopulation 96 stderr 1.598 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_aime.C3_jItdR.json) Source Vals AI Original model label openai/gpt-4o-2024-11-20 Source version 1 Observed 2026-10-21T16:46:01.920Z Evidence SHA-256 `7d016210b39d24351bbb5d12be09a973decac461bb2486eb1f23724fed2b7e6c` Source locator `/tasks/aime_2025/openai~1gpt-4o-2024-11-20` |  |  |  |  |  |  |  |
| 35 / 35 | [Claude 3.5 Sonnet](https://benchize.com/models/claude-3-5-sonnet.md) | 3.33% | 95% CI 0.12–6.54% | 0.009usd per problem | — | [matharena · aime--aime_2025](https://matharena.ai/competition_tables/aime--aime_2025) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 3.33 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.009 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition aime--aime_2025 predicted false sourceModelName Claude-3.5-Sonnet sourceModelVersion anthropic/claude-35 ### Reported statistics Uncertainty 95% confidence interval: 0.12–6.54% confidenceInterval95 3.21 problems 30 ### Source [Original source](https://matharena.ai/competition_tables/aime--aime_2025) Source matharena Original model label Claude-3.5-Sonnet Source version aime--aime_2025 Observed 2026-10-21T16:52:35.082Z Evidence SHA-256 `fab0abcecc4032852ef2cbd2a84d34ffc42202be358b2994bfaba44ae7a0cd9a` Source locator `/table/rows/60` |  |  |  |  |  |  |  |

---
Cite as: AIME 2025 — Benchize, https://benchize.com/benchmarks/aime-2025 (retrieved 2026-10-21)