{"id":22290,"date":"2025-04-17T11:24:17","date_gmt":"2025-04-17T11:24:17","guid":{"rendered":"https:\/\/metaverseplanet.net\/blog\/?p=22290"},"modified":"2026-01-05T09:16:18","modified_gmt":"2026-01-05T09:16:18","slug":"openai-unveils-new-ai-models-o3-and-o4-mini","status":"publish","type":"post","link":"https:\/\/metaverseplanet.net\/blog\/openai-unveils-new-ai-models-o3-and-o4-mini\/","title":{"rendered":"OpenAI Unveils New AI Models: o3 and o4-mini"},"content":{"rendered":"\n

OpenAI has introduced two new AI models, o3<\/strong> and o4-mini<\/strong>. What sets these models apart is that, rather than responding immediately, they first pause to reflect\u2014engaging in an internal reasoning process to deliver more logical and accurate answers.<\/p>\n\n\n\n

The o3<\/strong> model emerges as OpenAI\u2019s most powerful \u201cthinking\u201d AI to date. It appears to outperform previous versions across a wide range of tasks\u2014from mathematics and coding to logical reasoning, scientific analysis, and visual interpretation<\/strong>. Meanwhile, the o4-mini<\/strong> model offers a more cost\u2011effective alternative, striking a balance between speed<\/strong> and performance<\/strong>. OpenAI has confirmed that both models can leverage the built\u2011in ChatGPT tools for web browsing<\/strong>, Python code execution<\/strong>, image processing<\/strong>, and image generation<\/strong>, and that they can incorporate analysis of user\u2011uploaded images directly into their response workflow.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

These new models are available immediately to subscribers of the ChatGPT Pro<\/strong>, Plus<\/strong>, and Team<\/strong> plans. In addition, OpenAI has rolled out an o4-mini-high<\/strong> variant, which allocates extra compute time to further boost response accuracy.<\/p>\n\n\n\n

In internal evaluations using the SWE\u2011bench test (which measures coding proficiency), o3<\/strong> achieved a score of 69.1%<\/strong>, while o4-mini<\/strong> scored 68.1%<\/strong>. Beyond their enhanced visual processing<\/strong> capabilities, both models can execute Python code directly in the browser and perform live internet searches<\/strong> to access the latest information.<\/p>\n\n\n\n

For developers, OpenAI offers integration via the Chat Completions API and the Responses API. Pricing is set at $10<\/strong> per million input tokens and $40<\/strong> per million output tokens for the o3<\/strong> model; the o4-mini<\/strong> model carries the same rates as its predecessor, o3-mini<\/strong>.<\/p>\n\n\n\n

Looking ahead, OpenAI plans to launch an o3-pro<\/strong> model exclusively for ChatGPT Pro subscribers in the coming weeks. CEO Sam Altman<\/strong> noted that o3<\/strong> and o4-mini<\/strong> may be the final standalone reasoning\u2011focused models released before GPT-5<\/strong>.<\/p>\n\n\n\n\n

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

    OpenAI has introduced two new AI models, o3 and o4-mini. What sets these models apart is that, rather than responding immediately, they first pause to reflect\u2014engaging in an internal reasoning process to deliver more logical and accurate answers. The o3 model emerges as OpenAI\u2019s most powerful \u201cthinking\u201d AI to date. It appears to outperform previous …<\/p>\n","protected":false},"author":1,"featured_media":17398,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"googlesitekit_rrm_CAown96uCw:productID":"","footnotes":""},"categories":[332],"tags":[335,210],"class_list":["post-22290","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-information","tag-ai-news","tag-ai-tools-news"],"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/22290","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/comments?post=22290"}],"version-history":[{"count":0,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/22290\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media\/17398"}],"wp:attachment":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media?parent=22290"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/categories?post=22290"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/tags?post=22290"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}