# BrokenArXiv 06/2026 > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/matharena%3Aarxiv_false--june. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/matharena%3Aarxiv_false--june. BrokenArXiv 06/2026: Recognizing false mathematical claims derived from recent research papers instead of producing invalid proofs. This competition contains 54 problems. MathArena reports normalized performance across repeated attempts and a 95% confidence interval. Canonical URL: https://benchize.com/benchmarks/matharena%3Aarxiv_false--june 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 BrokenArXiv 06/2026: Recognizing false mathematical claims derived from recent research papers instead of producing invalid proofs. This competition contains 54 problems. MathArena reports normalized performance across repeated attempts and a 95% confidence interval. Mathematics · Models tested: 21 · Score publishers: [matharena](https://matharena.ai/?comp=arxiv_false--june) ## Model results | # | Model | Score | ± | Cost | | | --- | --- | --- | --- | --- | --- | | 1 | [Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-high.md)High | 100% | 95% CI 100–100% | 0.06usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 100.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.060 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort high sourceModelName Claude-Opus-5.5 (high) sourceModelVersion anthropic/opus_55_high thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 100–100% confidenceInterval95 0.00 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Claude-Opus-5.5 (high) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/0` | | | | | | | 1 | [GPT-6.1 Sol](https://benchize.com/models/gpt-6-1-sol-max.md)Max | 100% | 95% CI 100–100% | 0.16usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 100.00 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.16 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName GPT-6.1 Sol (max) sourceModelVersion openai/gpt-61-sol thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 100–100% confidenceInterval95 0.00 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-6.1 Sol (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/1` | | | | | | | 3 | [GPT-6 Astra](https://benchize.com/models/gpt-6-astra-max.md)Max | 99.07% | 95% CI 97.26–100.88% | 0.63usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 99.07 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.63 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName GPT-6 Astra (max) sourceModelVersion openai/gpt-6-astra thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 97.26–100.88% confidenceInterval95 1.81 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-6 Astra (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/2` | | | | | | | 4 | [GPT-6 Astra](https://benchize.com/models/gpt-6-astra-low.md)Low | 97.69% | 95% CI 94.85–100.53% | 0.08usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 97.69 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.079 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort low sourceModelName GPT-6 Astra (low) sourceModelVersion openai/gpt-6-astra-low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 94.85–100.53% confidenceInterval95 2.84 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-6 Astra (low) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/3` | | | | | | | 5 | [GPT-6 Sol](https://benchize.com/models/gpt-6-sol-max.md)Max | 95.37% | 95% CI 91.41–99.33% | 0.29usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.37 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.29 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName GPT-6 Sol (max) sourceModelVersion openai/gpt-6-sol thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 91.41–99.33% confidenceInterval95 3.96 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-6 Sol (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/4` | | | | | | | 6 | [Claude Opus 5](https://benchize.com/models/claude-opus-5-max.md)Max | 90.74% | 95% CI 85.27–96.21% | 2.92usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.74 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 2.92 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName Claude-Opus-5 (max) sourceModelVersion anthropic/opus_5_max thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 85.27–96.21% confidenceInterval95 5.47 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Claude-Opus-5 (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/5` | | | | | | | 7 | [Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-low.md)Low | 86.57% | 95% CI 80.14–93% | 1.12usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.57 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 1.12 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort low sourceModelName Claude-Fable-5.1 (low) sourceModelVersion anthropic/fable_51_low thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 80.14–93% confidenceInterval95 6.43 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Claude-Fable-5.1 (low) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/6` | | | | | | | 8 | [GPT-5.5](https://benchize.com/models/gpt-5-5-xhigh.md)Extra high | 69.44% | 95% CI 62.35–76.53% | 1.47usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 69.44 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 1.47 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort xhigh sourceModelName GPT-5.5 (xhigh) sourceModelVersion openai/gpt-55 thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 62.35–76.53% confidenceInterval95 7.09 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-5.5 (xhigh) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/7` | | | | | | | 9 | [GPT-5.6 Sol](https://benchize.com/models/gpt-5-6-sol-max.md)Max | 67.28% | 95% CI 60.06–74.5% | — | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 67.28 \| percent \| competition \| Higher is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName GPT-5.6-Sol (max) sourceModelVersion openai/gpt-56-sol thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 60.06–74.5% confidenceInterval95 7.22 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GPT-5.6-Sol (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/8` | | | | | | | 10 | [Claude Fable 5](https://benchize.com/models/claude-fable-5-max.md)Max | 47.84% | 95% CI 40.15–55.53% | 6.21usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 47.84 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 6.21 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName Claude-Fable-5 (max) sourceModelVersion anthropic/fable_5_max thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 40.15–55.53% confidenceInterval95 7.69 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Claude-Fable-5 (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/13` | | | | | | | 11 | [Qwen3.8 Max](https://benchize.com/models/qwen3-8-max.md) | 43.98% | 95% CI 34.62–53.34% | 0.38usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 43.98 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.38 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Qwen3.8-Max sourceModelVersion qwen/qwen3.8_max ### Reported statistics Uncertainty 95% confidence interval: 34.62–53.34% confidenceInterval95 9.36 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Qwen3.8-Max Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/14` | | | | | | | 12 | [Claude Opus 4.8](https://benchize.com/models/claude-opus-4-8-max.md)Max | 41.67% | 95% CI 34.08–49.26% | 5.55usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 41.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 5.55 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName Claude-Opus-4.8 (max) sourceModelVersion anthropic/opus_48_max thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 34.08–49.26% confidenceInterval95 7.59 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Claude-Opus-4.8 (max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/16` | | | | | | | 13 | [DeepSeek V4.1 Flash](https://benchize.com/models/deepseek-v4-1-flash-max.md)Max | 39.35% | 95% CI 30.14–48.56% | 0.03usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 39.35 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.034 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName DeepSeek-V4.1-Flash (Max) sourceModelVersion deepseek/deepseek_v41_flash thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 30.14–48.56% confidenceInterval95 9.21 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label DeepSeek-V4.1-Flash (Max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/17` | | | | | | | 14 | [Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash.md) | 20.83% | 95% CI 13.17–28.49% | 0.16usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 20.83 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.16 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Gemini 3.8 Flash sourceModelVersion gemini/gemini-38-flash ### Reported statistics Uncertainty 95% confidence interval: 13.17–28.49% confidenceInterval95 7.66 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Gemini 3.8 Flash Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/20` | | | | | | | 15 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview.md) | 17.59% | 95% CI 11.73–23.45% | 0.33usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 17.59 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.33 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Gemini 3.1 Pro Preview sourceModelVersion gemini/gemini-31-pro ### Reported statistics Uncertainty 95% confidence interval: 11.73–23.45% confidenceInterval95 5.86 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Gemini 3.1 Pro Preview Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/21` | | | | | | | 16 | [DeepSeek V4 Flash](https://benchize.com/models/deepseek-v4-flash-max.md)Max | 16.67% | 95% CI 10.93–22.41% | 0.04usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 16.67 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.037 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false Reasoning effort max sourceModelName DeepSeek-v4-Flash (Max) sourceModelVersion deepseek/deepseek_v4_flash thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 10.93–22.41% confidenceInterval95 5.74 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label DeepSeek-v4-Flash (Max) Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/22` | | | | | | | 17 | [Step 3.7 Flash](https://benchize.com/models/step-3-7-flash.md) | 15.74% | 95% CI 10.13–21.35% | 0.03usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 15.74 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.027 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Step 3.7 Flash sourceModelVersion stepfun/3.7-flash ### Reported statistics Uncertainty 95% confidence interval: 10.13–21.35% confidenceInterval95 5.61 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Step 3.7 Flash Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/23` | | | | | | | 18 | [GLM-5.2](https://benchize.com/models/glm-5-2.md) | 14.2% | 95% CI 8.83–19.57% | 0.14usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 14.20 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.14 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName GLM 5.2 sourceModelVersion glm/glm-52 ### Reported statistics Uncertainty 95% confidence interval: 8.83–19.57% confidenceInterval95 5.37 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label GLM 5.2 Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/24` | | | | | | | 19 | [Gemini 3.5 Flash](https://benchize.com/models/gemini-3-5-flash.md) | 12.35% | 95% CI 7.28–17.42% | 0.32usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 12.35 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.32 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Gemini 3.5 Flash sourceModelVersion gemini/gemini-35-flash ### Reported statistics Uncertainty 95% confidence interval: 7.28–17.42% confidenceInterval95 5.07 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Gemini 3.5 Flash Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/25` | | | | | | | 20 | [Gemini 3.6 Flash](https://benchize.com/models/gemini-3-6-flash.md) | 12.04% | 95% CI 5.9–18.18% | 0.15usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 12.04 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.15 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Gemini 3.6 Flash sourceModelVersion gemini/gemini-36-flash ### Reported statistics Uncertainty 95% confidence interval: 5.9–18.18% confidenceInterval95 6.14 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Gemini 3.6 Flash Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/26` | | | | | | | 21 | [Gemini 3.7 Flash](https://benchize.com/models/gemini-3-7-flash.md) | 11.11% | 95% CI 5.18–17.04% | 0.08usd per problem | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 11.11 \| percent \| competition \| Higher is better \| \| `cost_per_problem` \| 0.078 \| usd_per_problem \| competition \| Lower is better \| ### Configuration competition arxiv_false--june predicted false sourceModelName Gemini 3.7 Flash sourceModelVersion gemini/gemini-37-flash ### Reported statistics Uncertainty 95% confidence interval: 5.18–17.04% confidenceInterval95 5.93 problems 54 ### Source [Original source](https://matharena.ai/competition_tables/arxiv_false--june) Source matharena Original model label Gemini 3.7 Flash Source version arxiv_false--june Observed 2026-10-21T16:52:25.323Z Evidence SHA-256 `f6d7e2280d05706643d46ae7786bd3fad6d912e4893781c3f62c73deebf38459` Source locator `/table/rows/27` | | | | | | --- Cite as: BrokenArXiv 06/2026 — Benchize, https://benchize.com/benchmarks/matharena%3Aarxiv_false--june (retrieved 2026-10-21)