{"id":1458,"date":"2025-07-16T21:06:00","date_gmt":"2025-07-16T13:06:00","guid":{"rendered":"http:\/\/www.applicationsofllm.com\/?p=1458"},"modified":"2025-09-17T21:09:17","modified_gmt":"2025-09-17T13:09:17","slug":"%e3%80%90agi-eval%e8%af%84%e6%b5%8b%e6%8a%a5%e5%91%8a-no-6%e3%80%91o3-o4-mini-%e6%96%87%e6%9c%ac%e6%9d%83%e5%a8%81%e8%af%84%e6%b5%8b%ef%bc%9ao3-%e5%bc%ba%e5%8a%bf%e7%99%bb%e9%a1%b6","status":"publish","type":"post","link":"http:\/\/www.applicationsofllm.com\/index.php\/2025\/07\/16\/%e3%80%90agi-eval%e8%af%84%e6%b5%8b%e6%8a%a5%e5%91%8a-no-6%e3%80%91o3-o4-mini-%e6%96%87%e6%9c%ac%e6%9d%83%e5%a8%81%e8%af%84%e6%b5%8b%ef%bc%9ao3-%e5%bc%ba%e5%8a%bf%e7%99%bb%e9%a1%b6\/","title":{"rendered":"\u3010AGI-Eval\u8bc4\u6d4b\u62a5\u544a NO.6\u3011o3 \/ o4-mini \u6587\u672c\u6743\u5a01\u8bc4\u6d4b\uff1ao3 \u5f3a\u52bf\u767b\u9876"},"content":{"rendered":"\n<p>o3 \u548c o4 mini \u7684\u5168\u65b9\u4f4d\u8bc4\u6d4b\u7ed3\u679c\u6765\u5566\uff01o3&nbsp;\u76f4\u63a5\u5728\u6587\u672c\u63a8\u7406\u65b9\u5411\u4e0a\u767b\u9876\uff01\u5728\u4ea4\u4e92\u80fd\u529b\u548c\u6307\u4ee4\u9075\u5faa\u7b49\u65b9\u9762\u5904\u4e8e\u9886\u5148\u5730\u4f4d\uff0c\u4f46\u5728\u77e5\u8bc6\u50a8\u5907\u65b9\u5411\u7a0d\u663e\u4e0d\u8db3\u3002o4 mini \u4e5f\u8868\u73b0\u4eae\u773c\uff0c\u5728\u4ee3\u7801\u65b9\u5411\uff0c\u5f3a\u4e8e OpenAI-o3-mini (high)\u3001OpenAI-o3\u548cDeepseek-R1 \u7b49\u4e00\u4f17\u6a21\u578b\uff0c\u51c6\u786e\u7387\u767b\u9876\u7b2c\u4e00\uff0c\u5728\u63a8\u7406\u65b9\u5411\u5904\u4e8e\u5c0f\u578b\u63a8\u7406\u6a21\u578b Top1\u3002\u90a3\u4e48\u4e00\u8d77\u6765\u770b\u4e0b\u672c\u6b21\u8bc4\u6d4b\u7ed3\u679c\uff01<\/p>\n\n\n\n<p>\u76ee\u5f55\uff1a<\/p>\n\n\n\n<p>1.&nbsp;o3 \u548c o4 mini\u8bc4\u6d4b\u5206\u6790<\/p>\n\n\n\n<p>&nbsp; &nbsp;1.1 \u8bc4\u6d4b\u6982\u8ff0<\/p>\n\n\n\n<p>&nbsp; &nbsp;1.2 o3 \u8bc4\u6d4b\u7ed3\u679c<\/p>\n\n\n\n<p>&nbsp; &nbsp;1.3 o4-mini\u8bc4\u6d4b\u7ed3\u679c<\/p>\n\n\n\n<p>2.&nbsp;o3 \u548c o4 mini \u4fe1\u606f\u56de\u987e<\/p>\n\n\n\n<p><strong>1.o3 \u548c o4 mini\u8bc4\u6d4b\u5206\u6790<\/strong><\/p>\n\n\n\n<p><strong>1.1 \u8bc4\u6d4b\u6982\u8ff0<\/strong><\/p>\n\n\n\n<p>\u3010\u8bc4\u6d4b\u5bf9\u8c61\u3011<\/p>\n\n\n\n<p>OpenAI-o3 \u4e0e OpenAI-o4-mini<\/p>\n\n\n\n<p>\u3010\u8bc4\u6d4b\u65b9\u5f0f\u3011<\/p>\n\n\n\n<p>\u57fa\u4e8e\u81ea\u5efa\u7684\u901a\u7528\u80fd\u529b\u8bc4\u6d4b\u96c6\u4e0e\u516c\u5f00\u8bc4\u6d4b\u96c6<\/p>\n\n\n\n<p>\u3010\u8bc4\u6d4b\u7ed3\u8bba\u3011<\/p>\n\n\n\n<p><strong>\u2460 o3 \u6beb\u65e0\u60ac\u5ff5\u7684\u767b\u9876\uff0c\u5728\u4ea4\u4e92\u80fd\u529b\u3001\u63a8\u7406\u80fd\u529b\u3001\u6307\u4ee4\u9075\u5faa\u548c\u521d\u7b49\u6570\u5b66\u56db\u4e2a\u65b9\u9762\u5168\u9762\u9886\u5148\uff1b<\/strong><\/p>\n\n\n\n<p><strong>\u2461&nbsp;o4-mini&nbsp;\u4ee3\u7801\u51c6\u786e\u7387\u767b\u9876 Top1\uff0c\u9ad8\u8fbe0.5400\uff0c\u6a21\u578b\u6392\u540d\u4e5f\u6765\u5230\u4e86\u7b2c\u4e09\u3002<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"657\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-218-1024x657.png\" alt=\"\" class=\"wp-image-1461\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-218-1024x657.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-218-300x193.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-218-768x493.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-218.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u63a5\u4e0b\u6765\u6211\u4eec\u7ee7\u7eed\u4e3a\u5927\u5bb6\u62c6\u89e3 o3 \u548c o4-mini \u7684\u8be6\u7ec6\u8bc4\u6d4b\u5185\u5bb9<\/p>\n\n\n\n<p><strong>1.2 &nbsp;o3 \u8bc4\u6d4b\u7ed3\u679c<\/strong><\/p>\n\n\n\n<p><strong>o3 \u6838\u5fc3\u7ed3\u8bba\uff1a<\/strong><\/p>\n\n\n\n<p>o3 \u5728\u7efc\u5408\u80fd\u529b\u4e0a\u8868\u73b0\u4f18\u5f02\uff0c\u5c24\u5176\u5728\u4ea4\u4e92\u80fd\u529b\u3001\u63a8\u7406\u80fd\u529b\u548c\u6307\u4ee4\u9075\u5faa\u65b9\u9762\u5904\u4e8e\u9886\u5148\u5730\u4f4d\u3002\u4f46\u662f\u5176\u77e5\u8bc6\u50a8\u5907\u76f8\u8f83\u4e8e OpenAI-o1 \u7a0d\u663e\u4e0d\u8db3\uff0c\u540c\u65f6\u5728\u8ba4\u77e5\u80fd\u529b\u65b9\u9762\u4e0e DeepSeek-R1 \u5b58\u5728\u4e00\u5b9a\u5dee\u8ddd\u3002<\/p>\n\n\n\n<p><strong>\u540c\u7c7b\u578b\u6a21\u578b\u80fd\u529b\u5bf9\u6bd4\uff1a<\/strong><\/p>\n\n\n\n<p>\u672c\u6b21\u8bc4\u6d4b\uff0c\u6211\u4eec\u57fa\u4e8e\u81ea\u5efa\u7684\u901a\u7528\u80fd\u529b\u8bc4\u6d4b\u96c6\u4e0e\u516c\u5f00\u8bc4\u6d4b\u96c6\uff0c\u5c06 OpenAI o3 \u4e0e o1 \u8fdb\u884c\u4e86\u5bf9\u6bd4\uff1a<\/p>\n\n\n\n<p><strong>\u5728\u901a\u7528\u80fd\u529b\u4e0a<\/strong>\uff0cOpenAI o3 \u6c34\u4f4d\u4f18\u4e8e OpenAI o1\uff0c\u5176\u4e2d<strong>\u63a8\u7406\u80fd\u529b\u6709\u660e\u663e\u63d0\u5347\uff0c\u4f46\u8ba4\u77e5\u80fd\u529b\u7a0d\u6709\u4e0b\u964d<\/strong>\uff1b<\/p>\n\n\n\n<p><strong>\u5728\u8f93\u51fa\u957f\u5ea6\u4e0a<\/strong>\uff0cOpenAI o3 \u5728\u4ee3\u7801\u63a8\u7406\u80fd\u529b\u63d0\u5347\u7684\u540c\u65f6\uff0c\u8f93\u51fa\u957f\u5ea6\u6ca1\u6709\u589e\u52a0\uff0c\u53cd\u800c\u51cf\u5c0f\u4e86\uff0c\u6ca1\u6709\u901a\u8fc7\u8f93\u51fa\u6210\u672c\u7f6e\u6362\u80fd\u529b\u7684\u63d0\u5347\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"361\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-219-1024x361.png\" alt=\"\" class=\"wp-image-1462\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-219-1024x361.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-219-300x106.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-219-768x271.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-219.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>o3 \u5728\u63a8\u7406\u6a21\u578b\u9886\u57df\u5904\u4e8e\u9886\u5148\u5730\u4f4d\uff0c\u5176\u80dc\u7387\u5747\u663e\u8457\u4f18\u4e8e\u540c\u7cfb\u5217\u4ea7\u54c1 OpenAI-o1\uff0c\u63a8\u7406\u80fd\u529b\u63d0\u5347\u8f83\u4e3a\u663e\u8457\u3002<\/p>\n\n\n\n<p><strong>1.3 o4-mini \u8bc4\u6d4b\u7ed3\u679c<\/strong><\/p>\n\n\n\n<p><strong>o4-mini \u6838\u5fc3\u7ed3\u8bba\uff1a<\/strong><\/p>\n\n\n\n<p><strong>o4-mini&nbsp;\u5904\u4e8e\u63a8\u7406\u6a21\u578b\u7b2c\u4e00\u68af\u961f<\/strong>\uff0c\u548c o1 \u80fd\u529b\u76f8\u5f53\uff0c\u4f46\u76f8\u8f83 o3 \u4ecd\u6709\u5dee\u8ddd\uff1bo4-mini&nbsp;\u5904\u4e8e\u5c0f\u578b\u63a8\u7406\u6a21\u578b Top1\u3002<\/p>\n\n\n\n<p><strong>\u540c\u7c7b\u578b\u6a21\u578b\u80fd\u529b\u5bf9\u6bd4\uff1a<\/strong><\/p>\n\n\n\n<p>\u6211\u4eec\u5c06\u540c\u4e3a\u5c0f\u578b\u63a8\u7406\u6a21\u578b\u7684 OpenAI-o4-mini \u4e0e&nbsp;o3-mini \u8fdb\u884c\u4e86\u5bf9\u6bd4\uff1a<\/p>\n\n\n\n<p><strong>\u5728\u901a\u7528\u80fd\u529b\u4e0a<\/strong>\uff0co4-mini&nbsp;\u6c34\u5e73\u4f18\u4e8e o3-mini\uff0c<strong>\u5176\u4e2d\u63a8\u7406\u80fd\u529b\u63d0\u5347\u8f83\u4e3a\u663e\u8457\uff0c\u4ec5\u6b21\u4e8e Top1\uff08o3\uff09<\/strong>\uff1b\u4f46\u6307\u4ee4\u9075\u5faa\u5f31\u4e8e o3-mini\uff08medium\/high\uff09\uff1b<\/p>\n\n\n\n<p><strong>\u5728\u4ea4\u4e92\u80fd\u529b\u4e0a<\/strong>\uff0co4-mini&nbsp;\u8f83 o3-mini\uff08high\uff09\u6709\u63d0\u5347<strong>\uff0c<\/strong>\u4f46\u8ddd\u79bb o3\u3001o1 \u4ecd\u6709\u5dee\u8ddd\uff0c\u4e14\u843d\u540e\u4e8e DeepseekR1 \u548c V3\u3001doubao-1.5-pro\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"360\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-217-1024x360.png\" alt=\"\" class=\"wp-image-1459\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-217-1024x360.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-217-300x106.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-217-768x270.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-217.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>o4-mini&nbsp;\u4ee3\u7801\u51c6\u786e\u7387\u767b\u9876 Top1\uff0c\u5f3a\u4e8e OpenAI-o3-mini\uff08high\uff09\uff0cOpenAI-o3\uff0cDeepseek-R1\u3002<\/p>\n\n\n\n<p>\u3010\u6307\u6807\u6ce8\u91ca\u8bf4\u660e\u3011<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u5ba2\u89c2\u8bc4\u6d4b\u51c6\u786e\u7387\u6307\u6807\u8bf4\u660e\uff1a\u6a21\u578b\u56de\u7b54\u6b63\u786e\u6570\u91cf\/\u603b\u8bc4\u6d4b\u6570\u636e\u603b\u91cf\uff1b<\/li>\n\n\n\n<li>\u4e3b\u89c2\u8bc4\u6d4b\u80dc\u7387\u6307\u6807\u8bf4\u660e\uff1a(1*N_\u660e\u663e\u597d+0.75*N_\u7a0d\u597d+0.5*N_\u4e8c\u8005\u5dee\u4e0d\u591a+0.25*N_\u7a0d\u5dee+0*N_\u660e\u663e\u5dee)\/\u603b\u8bc4\u6d4b\u6570\u636e\u91cf<\/li>\n<\/ul>\n\n\n\n<p>* \u8bc4\u6d4b\u6307\u6807\u662f\u88ab\u8bc4\u6d4b\u6a21\u578b\u76f8\u5bf9\u4e8e\u6807\u6746\u6a21\u578b\u7684\u4e94\u6863\u52a0\u6743\u80dc\u7387<\/p>\n\n\n\n<p>*\u6807\u6746\u6a21\u578b\u4e3a GPT4-turbo-0125<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u8bc4\u6d4b\u65b9\u6848\uff1a\u53ef\u70b9\u51fb<strong>\u9605\u8bfb\u539f\u6587<\/strong>\u8df3\u8f6c\u5b98\u7f51\uff0c\u8bc4\u6d4b\u699c\u5355\u5904\u70b9\u51fb\u8bc4\u6d4b\u65b9\u6848\u67e5\u770b<\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"532\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-220-1024x532.png\" alt=\"\" class=\"wp-image-1463\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-220-1024x532.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-220-300x156.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-220-768x399.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-220.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p><strong>2.o3 \u548c o4 mini \u5b9e\u6d4b\u56de\u987e<\/strong><\/p>\n\n\n\n<p>\u4e0a\u4e00\u671f\u6211\u4eec\u5bf9 o3 \u548c o4 mini \u8fdb\u884c\u4e86\u5b9e\u6d4b\uff0c\u51fa\u73b0\u4e86\u4e0d\u5c11\u7ffb\u8f66\u6848\u4f8b\uff0c\u65e0\u6cd5\u8bc6\u522b\u6897\u56fe\u3001\u627e\u56fe\u7247\u4e0d\u540c\u4e4b\u5904\u627e\u4e0d\u9f50\u4e14\u8bc6\u522b\u6709\u9519\u8bef\u7b49\u7b49\uff0c\u8be6\u89c1<strong><a href=\"https:\/\/mp.weixin.qq.com\/s?__biz=Mzk2NDE2MjMxOA==&amp;mid=2247484974&amp;idx=1&amp;sn=e24fe5a8e348144f32689ce6fed0923b&amp;scene=21#wechat_redirect\" target=\"_blank\" rel=\"noreferrer noopener\">\u3010AGI-Eval \u5b9e\u6d4b\u901f\u62a5\u3011OpenAI o3\/o4-mini\u89c6\u89c9\u63a8\u7406\u5c01\u795e\uff1f\u5b9e\u6d4b\u7ffb\u8f66\u7387\u7565\u9ad8<\/a><\/strong><\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"689\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-221-1024x689.png\" alt=\"\" class=\"wp-image-1464\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-221-1024x689.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-221-300x202.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-221-768x517.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-221.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"564\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-222-1024x564.png\" alt=\"\" class=\"wp-image-1465\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-222-1024x564.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-222-300x165.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-222-768x423.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-222.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>\u867d\u7136&nbsp;o3 \u548c o4 mini \u5728\u5b9e\u6d4b\u4e2d\u6709\u4e0d\u5c11\u7ffb\u8f66\u6848\u4f8b\uff0c\u4f46\u662f\u5728\u66f4\u52a0\u5168\u9762\u3001\u79d1\u5b66\u7684\u8bc4\u6d4b\u4e2d\uff0c\u6211\u4eec\u5370\u8bc1\u4e86 OpenAI o3 \u662f OpenAI \u6700\u5f3a\u5927\u7684\u63a8\u7406\u6a21\u578b\uff0c\u5b83\u63a8\u52a8\u4e86\u7f16\u7801\u3001\u6570\u5b66\u3001\u79d1\u5b66\u3001\u89c6\u89c9\u611f\u77e5\u7b49\u9886\u57df\u7684\u53d1\u5c55\u3002\u5b83\u5728\u5305\u62ec Codeforces\u3001SWE-bench\uff08\u65e0\u9700\u6784\u5efa\u81ea\u5b9a\u4e49\u6a21\u578b\u7279\u5b9a\u652f\u67b6\uff09\u548c MMMU \u5728\u5185\u7684\u57fa\u51c6\u6d4b\u8bd5\u4e2d\u8bbe\u5b9a\u4e86\u65b0\u7684 SOTA\u3002<\/p>\n\n\n\n<p>\u5b83\u975e\u5e38<strong>\u9002\u5408\u9700\u8981\u591a\u65b9\u9762\u5206\u6790\u4e14\u7b54\u6848\u53ef\u80fd\u4e0d\u662f\u7acb\u5373\u663e\u800c\u6613\u89c1\u7684\u590d\u6742\u67e5\u8be2<\/strong>\uff0c\u5728\u5206\u6790\u56fe\u50cf\u3001\u56fe\u8868\u548c\u56fe\u5f62\u7b49\u89c6\u89c9\u4efb\u52a1\u4e2d\u8868\u73b0\u5c24\u4e3a\u51fa\u8272\u3002\u5728\u5916\u90e8\u4e13\u5bb6\u7684\u8bc4\u4f30\u4e2d\uff0co3 \u5728\u56f0\u96be\u7684\u73b0\u5b9e\u4efb\u52a1\u4e2d\u6bd4 OpenAI o1 \u72af\u7684\u91cd\u5927\u9519\u8bef\u5c11 20%\uff0c\u5c24\u5176\u662f\u5728\u7f16\u7a0b\u3001\u5546\u4e1a\u54a8\u8be2\u548c\u521b\u610f\u6784\u601d\u7b49\u9886\u57df\u8868\u73b0\u51fa\u8272\u3002<\/p>\n\n\n\n<p>o3 \u4f5c\u4e3a\u5f53\u524d\u6700\u5f3a\u7684\u56fe\u50cf\u63a8\u7406\u6a21\u578b\uff0c \u5728V*\u6d4b\u8bd5\u4e2d\uff0c\u89c6\u89c9\u63a8\u7406\u65b9\u6cd5\u8fbe\u5230\u4e86 95.7% \u7684\u51c6\u786e\u7387\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"735\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-223-1024x735.png\" alt=\"\" class=\"wp-image-1466\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-223-1024x735.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-223-300x215.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-223-768x551.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-223.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>OpenAI-o4-mini \u662f\u4e00\u79cd\u8f83\u5c0f\u7684\u6a21\u578b\uff0c<strong>\u9488\u5bf9\u5feb\u901f\u3001\u7ecf\u6d4e\u9ad8\u6548\u7684\u63a8\u7406\u8fdb\u884c\u4e86\u4f18\u5316<\/strong>\u2014\u2014\u5b83\u5728\u5176\u5c3a\u5bf8\u548c\u6210\u672c\u65b9\u9762\u5b9e\u73b0\u4e86\u5353\u8d8a\u7684\u6027\u80fd\uff0c\u7279\u522b\u662f\u5728\u6570\u5b66\u3001\u7f16\u7801\u548c\u89c6\u89c9\u4efb\u52a1\u65b9\u9762\u3002\u5b83\u662f AIME 2024 \u548c 2025 \u4e0a\u8868\u73b0\u6700\u4f73\u7684\u57fa\u51c6\u6a21\u578b\u3002\u5728\u4e13\u5bb6\u8bc4\u4f30\u4e2d\uff0co4-mini \u5728\u975e STEM \u4efb\u52a1\u4ee5\u53ca\u6570\u636e\u79d1\u5b66\u7b49\u9886\u57df\u7684\u8868\u73b0\u5747\u4f18\u4e8e\u5176\u524d\u8eab o3-mini\u3002\u5f97\u76ca\u4e8e\u5176\u9ad8\u6548\u6027\uff0co4-mini \u652f\u6301\u7684\u4f7f\u7528\u9650\u5236\u8fdc\u9ad8\u4e8e o3\uff0c\u4f7f\u5176\u6210\u4e3a\u5904\u7406\u9700\u8981\u63a8\u7406\u80fd\u529b\u7684\u95ee\u9898\u7684\u5f3a\u5927\u9ad8\u5bb9\u91cf\u3001\u9ad8\u541e\u5410\u91cf\u89e3\u51b3\u65b9\u6848\u3002<\/p>\n\n\n\n<p>o4-mini\u63a8\u7406\u80fd\u529b\u589e\u5f3a\u4f34\u968f\u6210\u672c\u589e\u52a0\u3002\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"618\" src=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-224-1024x618.png\" alt=\"\" class=\"wp-image-1467\" srcset=\"http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-224-1024x618.png 1024w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-224-300x181.png 300w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-224-768x464.png 768w, http:\/\/www.applicationsofllm.com\/wp-content\/uploads\/2025\/09\/image-224.png 1108w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n","protected":false},"excerpt":{"rendered":"<p>o3 \u548c o4 mini \u7684\u5168\u65b9\u4f4d\u8bc4\u6d4b\u7ed3\u679c\u6765\u5566\uff01o3&nbsp;\u76f4\u63a5\u5728\u6587\u672c\u63a8\u7406\u65b9\u5411\u4e0a\u767b\u9876\uff01\u5728\u4ea4\u4e92\u80fd\u529b\u548c\u6307\u4ee4\u9075\u5faa [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":1460,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8],"tags":[],"class_list":["post-1458","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-8"],"_links":{"self":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1458","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/comments?post=1458"}],"version-history":[{"count":1,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1458\/revisions"}],"predecessor-version":[{"id":1468,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1458\/revisions\/1468"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media\/1460"}],"wp:attachment":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media?parent=1458"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/categories?post=1458"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/tags?post=1458"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}