{"id":10807,"date":"2024-05-21T09:49:25","date_gmt":"2024-05-21T01:49:25","guid":{"rendered":"https:\/\/www.1ai.net\/?p=10807"},"modified":"2024-05-21T09:49:25","modified_gmt":"2024-05-21T01:49:25","slug":"gpt-4o-%e5%8f%af%e4%bb%a5%e5%85%8d%e8%b4%b9%e4%bd%bf%e7%94%a8%e5%90%97%ef%bc%9fopenai-%e5%85%a8%e8%83%bd%e6%a8%a1%e5%9e%8bgpt-4o-%e6%80%8e%e4%b9%88%e4%bd%bf%e7%94%a8%ef%bc%9f","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/10807.html","title":{"rendered":"Can GPT-4o be used for free? How to use OpenAI's all-purpose model GPT-4o?"},"content":{"rendered":"

\u4e0a\u5468\uff0cOpenAI<\/a> \u53d1\u5e03\u4e86 GPT-4o<\/a> \u6a21\u578b\uff0c\u8fd9\u662f\u5176\u591a\u5e74\u6765\u5f00\u53d1\u7684 GPT-n \u7cfb\u5217\u6a21\u578b\u4e2d\u6700\u4ee4\u4eba\u5174\u594b\u7684\u6700\u65b0\u6a21\u578b\u3002<\/p>\n

OpenAI\u7684GPT-4o\u6a21\u578b\u5728\u591a\u529f\u80fd\u6027\u3001\u6027\u80fd\u548c\u54cd\u5e94\u901f\u5ea6\u65b9\u9762\u5747\u8d85\u8d8a\u4e86\u5176\u524d\u8eabGPT-4\u548cGPT-4 Turbo\u6a21\u578b\u3002<\/p>\n

\u6b64\u5916\uff0cGPT-4o \u6a21\u578b\u8fd8\u80fd\u5904\u7406\u7528\u6237\u8f93\u5165\u7684\u97f3\u9891\u6216\u56fe\u50cf\uff0c\u7136\u540e\u5feb\u901f\u5730\u751f\u6210\u8bed\u97f3\u6216\u6587\u672c\u8f93\u51fa\u3002\u5982\u679c\u4f60\u60f3\u6df1\u5165\u4e86\u89e3GPT-4o\u6a21\u578b\uff0c\u90a3\u4f60\u6765\u5bf9\u5730\u65b9\u4e86\uff01<\/p>\n

\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4f1a\u8be6\u7ec6\u4ecb\u7ecd GPT-4o \u6a21\u578b\uff0c\u4f60\u51c6\u5907\u597d\u4e86\u5417\uff1f<\/p>\n

\"get-411\"<\/div>\n

\u4ec0\u4e48\u662f GPT-4o\uff1f<\/strong><\/p>\n

GPT-4o \u662f OpenAI \u5f00\u53d1\u7684\u6700\u65b0\u4eba\u5de5\u667a\u80fd\u6a21\u578b\uff0c\u5b83\u62e5\u6709\u521b\u65b0\u7684\u591a\u6a21\u6001\u529f\u80fd\uff0c\u80fd\u591f\u5904\u7406\u548c\u7406\u89e3\u591a\u79cd\u683c\u5f0f\u7684\u4fe1\u606f\u3002\u4e0d\u4ec5\u53ef\u4ee5\u5904\u7406\u6587\u672c\uff0c\u8fdb\u884c\u5bf9\u8bdd\u548c\u521b\u4f5c\uff0cGPT-4o \u8fd8\u80fd\u5206\u6790\u97f3\u9891\u548c\u56fe\u50cf\u3002<\/p>\n

\u4f8b\u5982\uff0c\u5b83\u80fd\u591f\u542c\u6b4c\u66f2\u5e76\u5206\u6790\u5176\u60c5\u611f\uff0c\u6216\u67e5\u770b\u56fe\u7247\u5e76\u63cf\u8ff0\u573a\u666f\u3002\u8fd9\u79cd\u80fd\u529b\u8ba9 GPT-4o \u4e0d\u4ec5\u80fd\u7406\u89e3\u5b57\u9762\u610f\u4e49\u4e0a\u7684\u4fe1\u606f\uff0c\u8fd8\u80fd\u6355\u6349\u5230\u6c9f\u901a\u4e2d\u7684\u5fae\u5999\u5dee\u522b\uff0c\u4f7f\u5f97\u5bf9\u8bdd\u66f4\u52a0\u81ea\u7136\u548c\u5bcc\u6709\u5438\u5f15\u529b\u3002<\/p>\n

\u6b64\u5916\uff0cGPT-4o \u7684\u591a\u6a21\u5f0f\u7406\u89e3\u5927\u5927\u589e\u5f3a\u4e86\u4fe1\u606f\u5904\u7406\u80fd\u529b\uff0c\u4f7f\u5176\u80fd\u591f\u7efc\u5408\u6587\u672c\u3001\u97f3\u9891\u548c\u89c6\u89c9\u4fe1\u606f\u8fdb\u884c\u5206\u6790\uff0c\u4e3a\u5f00\u53d1\u65b0\u5e94\u7528\u5982\u4eba\u5de5\u667a\u80fd\u52a9\u624b\u3001\u6559\u80b2\u5de5\u5177\u548c\u521b\u610f\u5185\u5bb9\u751f\u6210\u7b49\u63d0\u4f9b\u4e86\u53ef\u80fd\u3002<\/p>\n

\u8fd9\u79cd\u6280\u672f\u7684\u8fdb\u6b65\u4e0d\u4ec5\u4ec5\u662f AI \u9886\u57df\u7684\u4e00\u4e2a\u7a81\u7834\uff0c\u66f4\u662f\u4eba\u5de5\u667a\u80fd\u4e0e\u4e16\u754c\u4e92\u52a8\u548c\u7406\u89e3\u65b9\u5f0f\u5411\u66f4\u63a5\u8fd1\u4eba\u7c7b\u65b9\u5f0f\u8fc8\u8fdb\u7684\u91cd\u5927\u98de\u8dc3\u3002<\/p>\n

GPT-4o \u5982\u4f55\u5de5\u4f5c\uff1f<\/strong><\/p>\n

OpenAI \u7684 GPT-4o \u6a21\u578b\u548c GPT-4 \u5728\u5904\u7406\u97f3\u9891\u3001\u89c6\u89c9\u6216\u6587\u672c\u8f93\u5165\u7684\u65b9\u5f0f\u4e0a\u6709\u6240\u4e0d\u540c\u3002GPT-4 \u6a21\u578b\u662f\u901a\u8fc7\u591a\u4e2a\u795e\u7ecf\u7f51\u7edc\u5e76\u7ed3\u5408\u5b83\u4eec\u7684\u8f93\u51fa\u6765\u54cd\u5e94\u6587\u672c\u8f93\u5165\u548c\u97f3\u9891\u8f93\u51fa\u7684\u3002<\/p>\n

\u4f46\u662f\uff0cGPT-4o \u6a21\u578b\u53ea\u7528\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u5c31\u80fd\u5b8c\u6210\u6240\u6709\u5de5\u4f5c\u3002\u8fd9\u6837\uff0cGPT-4o \u5c31\u80fd\u7406\u89e3\u8f93\u5165\u4e2d\u7684\u97f3\u8c03\u3001\u8bc6\u522b\u591a\u4e2a\u8bf4\u8bdd\u8005\u3001\u7406\u89e3\u80cc\u666f\u566a\u97f3\uff0c\u5e76\u751f\u6210\u66f4\u81ea\u7136\u3001\u66f4\u5bcc\u6709\u60c5\u611f\u7684\u56de\u7b54\u3002<\/p>\n

\u5982\u4f55\u4f7f\u7528 GPT-4o\uff1f<\/strong><\/p>\n

OpenAI \u5ba3\u5e03 GPT-4o \u6a21\u578b\u514d\u8d39\u5411\u7528\u6237\u5f00\u653e\uff0c\u4ee5\u4fbf\u5728\u5168\u7403\u8303\u56f4\u5185\u63a8\u5e7f\u3002\u5982\u679c\u4f60\u6709 OpenAI \u7684\u8d26\u53f7\uff08\u8fd8\u6ca1\u6709\u8d26\u53f7\uff0c \u8bf7\u67e5\u770bChatGPT\u8be6\u7ec6\u6ce8\u518c\u6559\u7a0b\uff1a2024 \u5e74ChatGPT\u8be6\u7ec6\u6ce8\u518c\u6559\u7a0b\uff0c\u8d85\u5168\u56fe\u6587\u7248\u64cd\u4f5c\u6d41\u7a0b\uff01\uff09\uff0c\u5c31\u53ef\u4ee5\u767b\u5f55\u5230 ChatGPT \u7f51\u9875\u6216\u624b\u673a\u5e94\u7528\u7a0b\u5e8f\u514d\u8d39\u4f7f\u7528 GPT-4o\u3002<\/p>\n

\u4e0d\u8fc7\uff0cGPT-4o \u6a21\u578b\u5bf9\u514d\u8d39\u7528\u6237\u4f7f\u7528\u6709\u6b21\u6570\u9650\u5236\u3002\u5982\u679c\u4f60\u9700\u8981\u9891\u7e41\u5730\u4f7f\u7528\u8fd9\u4e2a\u6a21\u578b\uff0c\u53ef\u4ee5\u8003\u8651\u6bcf\u6708\u652f\u4ed8 20 \u7f8e\u5143\u5347\u7ea7\u5230 Plus \u4f1a\u5458\uff0c\u5c31\u53ef\u4ee5\u83b7\u5f975\u500d\u4e8e\u514d\u8d39\u7528\u6237\u7684\u4f7f\u7528\u6b21\u6570\u3002<\/p>\n

\"get-412\"<\/div>\n

GPT-4o API \u4ef7\u683c<\/strong><\/p>\n

\u5982\u679c\u4f60\u60f3\u628a GPT-4o \u6a21\u578b\u7528\u4f5c API\uff0c\u53ea\u9700\u652f\u4ed8 GPT-4 Turbo \u6a21\u578b\u4e00\u534a\u7684\u4ef7\u683c\u5373\u53ef\u3002GPT-4o \u6a21\u578b\u6bcf\u767e\u4e07\u4e2a\u8f93\u5165Token\u6536\u8d39 5 \u7f8e\u5143\uff0c\u6bcf\u767e\u4e07\u4e2a\u8f93\u51faToken\u6536\u8d39 15 \u7f8e\u5143\u3002<\/p>\n

\"get-410\"<\/div>\n

GPT-4o \u7279\u70b9<\/strong><\/p>\n

GPT-4o \u662f OpenAI \u6700\u65b0\u3001\u6700\u5148\u8fdb\u7684\u6a21\u578b\uff0c\u5b83\u4e3a\u5404\u79cd\u4ee4\u4eba\u5174\u594b\u7684\u4f7f\u7528\u573a\u666f\u548c\u65b0\u7684\u673a\u4f1a\u6253\u5f00\u4e86\u5927\u95e8\u3002\u8fd9\u4e2a\u6a21\u578b\u5177\u6709\u5148\u8fdb\u7684\u591a\u6a21\u6001\u529f\u80fd\uff0c\u6027\u80fd\u4e5f\u6bd4\u4e4b\u524d\u7684\u4ea7\u54c1\u66f4\u597d\u3002\u6211\u4eec\u6765\u8be6\u7ec6\u4e86\u89e3\u4e00\u4e0b GPT-4o \u7684\u529f\u80fd\u3002<\/p>\n

GPT-4o \u6027\u80fd<\/strong><\/p>\n

\u5728\u5404\u79cd\u6027\u80fd\u6d4b\u8bd5\u5bf9\u6bd4\u4e2d\uff0cGPT-4o \u90fd\u8868\u73b0\u5f97\u975e\u5e38\u597d\uff0c\u8d85\u8fc7\u4e86\u5b83\u7684\u524d\u4efb GPT-4\uff0c\u4ee5\u53ca Claude 3 Opus \u548c Gemini Pro 1.5\u3002\u5b83\u4e0d\u4ec5\u5e94\u7528\u8303\u56f4\u5e7f\uff0c\u8fd8\u80fd\u5b9e\u65f6\u5904\u7406\u6570\u636e\u5e76\u8f93\u51fa\u7ed3\u679c\uff0c\u5176\u6587\u672c\u8d28\u91cf\u4e5f\u5f97\u5230\u4e86\u5f88\u9ad8\u7684\u8bc4\u5206\u3002<\/p>\n

\u6839\u636e OpenAI \u7684\u6587\u7ae0\uff0c\u5728\u8861\u91cf\u8bed\u8a00\u7406\u89e3\u80fd\u529b\u7684 LLMU \u6d4b\u8bd5\u4e2d\uff0cGPT-4o \u5f97\u5206\u9ad8\u8fbe 88.7%\uff0c\u800c GPT-4 \u548c Claude 3 Opus \u7684\u5f97\u5206\u5206\u522b\u662f 86.6% \u548c 86.8%\u3002<\/p>\n

\u6b64\u5916\uff0c\u5728\u8bc4\u4f30\u8ba1\u7b97\u80fd\u529b\u7684 MATH \u6d4b\u8bd5\u4e2d\uff0cGPT-4o \u4ee5 76.6% \u7684\u9ad8\u5206\u9886\u5148\uff0c\u800c\u5728 GPQA \u548c HumanEval \u7684\u5f97\u5206\u5206\u522b\u662f 53.6% \u548c 90.2%\u3002<\/p>\n

\"get-414\"<\/div>\n

\u89c6\u89c9\u7406\u89e3\u80fd\u529b<\/strong><\/p>\n

GPT-4o \u6a21\u578b\u6700\u663e\u8457\u7684\u7279\u70b9\u4e4b\u4e00\u662f\u5176\u89c6\u89c9\u7406\u89e3\u80fd\u529b\u3002\u5b83\u53ef\u4ee5\u5b9e\u65f6\u5206\u6790\u89c6\u89c9\u3001\u89c6\u9891\u548c\u89c6\u9891\u901a\u8bdd\u6570\u636e\uff0c\u5e76\u6839\u636e\u5206\u6790\u51fa\u7684\u7ed3\u679c\uff0c\u751f\u6210\u7c7b\u4f3c\u4eba\u7c7b\u7684\u56de\u7b54\u3002<\/p>\n

\u6839\u636e OpenAI \u7684\u6570\u636e\u663e\u793a\uff0c\u65e0\u8bba\u662f\u5728\u7406\u89e3\u56fe\u7247\u3001\u56fe\u8868\u8fd8\u662f\u56fe\u5f62\u7684\u6d4b\u8bd5\u4e2d\uff0cGPT-4o \u7684\u8868\u73b0\u90fd\u8fdc\u8d85\u5176\u4ed6\u540c\u7c7b\u6a21\u578b\u548c\u5b83\u4e4b\u524d\u7684\u4efb\u4e00\u7248\u672c\u3002<\/p>\n

\"get-413\"<\/div>\n

\u9664\u4e86\u8fd9\u4e9b\u6570\u636e\uff0c\u5728\u5b9e\u9645\u4f7f\u7528\u4e2d\uff0c\u4f60\u53ef\u4ee5\u7528\u5b9e\u65f6\u56fe\u50cf\u5411 GPT-4o \u63d0\u95ee\u3002\u5b83\u4e0d\u4ec5\u80fd\u7406\u89e3\u4f60\u7684\u95ee\u9898\uff0c\u8fd8\u80fd\u5feb\u901f\u8f6c\u5316\u4e3a\u56fe\u50cf\uff0c\u7136\u540e\u7ed9\u51fa\u7b80\u6d01\u3001\u7c7b\u4f3c\u4eba\u7c7b\u7684\u56de\u7b54\u3002<\/p>\n

\u8bed\u97f3\/\u97f3\u9891\u5904\u7406<\/strong><\/p>\n

GPT-4o \u5728\u5904\u7406\u8bed\u97f3\u6216\u97f3\u9891\u65b9\u9762\u4e5f\u975e\u5e38\u5f3a\uff0c\u5176\u53cd\u5e94\u901f\u5ea6\u51e0\u4e4e\u548c\u4eba\u4e00\u6837\u3002\u901a\u5e38\u4eba\u7c7b\u56de\u5e94\u5bf9\u8bdd\u7684\u65f6\u95f4\u5927\u7ea6\u662f 250 \u6beb\u79d2\uff0c\u800c GPT-4o \u5927\u6982\u9700\u8981 320 \u6beb\u79d2\u6765\u5206\u6790\u8bed\u97f3\u5e76\u505a\u51fa\u56de\u5e94\u3002<\/p>\n

\u76f8\u6bd4\u4e4b\u4e0b\uff0c GPT-4 \u5219\u9700\u8981 5.4 \u79d2\uff0cGPT-3.5 \u9700\u8981 2.8 \u79d2\uff0cGPT-4o \u7684\u901f\u5ea6\u5feb\u4e86\u5f88\u591a\uff0c\u8ddfGPT-4o \u6a21\u578b\u5bf9\u8bdd\u51e0\u4e4e\u548c\u548c\u771f\u4eba\u5bf9\u8bdd\u4e00\u6837\u6d41\u7545\u3002<\/p>\n

\"get-415\"<\/div>\n

\u867d\u7136 GPT-4o \u76ee\u524d\u6bcf\u79cd\u8bed\u8a00\u90fd\u6709\u56fa\u5b9a\u7684\u8bed\u97f3\uff0c\u4f46 OpenAI \u5ba3\u5e03\u5c06\u5728\u672a\u6765\u51e0\u5468\u5185\u589e\u52a0\u66f4\u591a\u79cd\u7c7b\u7684\u8bed\u97f3\u3002\u5373\u4fbf\u5982\u6b64\uff0cGPT-4o \u73b0\u6709\u7684\u8bed\u97f3\u542c\u8d77\u6765\u5df2\u7ecf\u5f88\u50cf\u771f\u4eba\u4e86\uff0c\u80fd\u8868\u8fbe\u60c5\u611f\uff0c\u8bf4\u8bdd\u65f6\u6709\u81ea\u7136\u7684\u505c\u987f\uff0c\u800c\u4e14\u8bf4\u5f97\u5f88\u6d41\u7545\u3002<\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

Last week, OpenAI released the GPT-4o model, the most exciting of the GPT-n series that it has developed over the years. OpenAI's GPT-4o model went beyond its predecessor GPT-4 and GPT-4 Turbo models in terms of multifunctionality, performance and response speed. In addition, the GPT-4o model allows for the processing of audio or image inputd by the user and the quick generation of voice or text output. If you want to get to the GPT-4o model, you're in the right place! In this paper, I'll describe the GPT-4o model. Are you ready? What's GPT-4o? GPT-4o is the latest artificial intelligence model developed by OpenAI, which has many innovations<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[144],"tags":[2582,190],"collection":[],"class_list":["post-10807","post","type-post","status-publish","format-standard","hentry","category-baike","tag-gpt-4o","tag-openai"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/10807","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=10807"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/10807\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=10807"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=10807"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=10807"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=10807"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}