{"id":27896,"date":"2025-09-01T13:55:06","date_gmt":"2025-09-01T13:55:06","guid":{"rendered":"https:\/\/metaverseplanet.net\/blog\/?p=27896"},"modified":"2026-01-05T09:10:48","modified_gmt":"2026-01-05T09:10:48","slug":"openai-unveils-its-most-advanced-conversational-model-gpt-realtime","status":"publish","type":"post","link":"https:\/\/metaverseplanet.net\/blog\/openai-unveils-its-most-advanced-conversational-model-gpt-realtime\/","title":{"rendered":"OpenAI Unveils Its Most Advanced Conversational Model: gpt-realtime"},"content":{"rendered":"\n

OpenAI<\/strong> has introduced gpt-realtime<\/strong>, a new conversational model that better follows complex instructions, produces more natural-sounding voices, and is 20% cheaper<\/strong>. The model comes with two new voices.<\/p>\n\n\n\n

In October 2024, OpenAI<\/a><\/em><\/strong> introduced the Realtime API<\/strong>, which allows developers to create low-latency, multimodal experiences. Since then, thousands of developers have used this API to build natural-language conversation experiences into their applications. Now, a more advanced conversational model, gpt-realtime<\/strong>, has been announced, which includes enhanced features.<\/p>\n\n\n\n

\n\n\n\n

More Advanced and Cheaper<\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

The new model is better at following complex instructions<\/strong>, and the error rate in tasks like hailing a ride has been reduced. The voices generated by the model are also noted to be more natural and expressive<\/strong>. OpenAI<\/strong> states that the model can interpret system messages and developer commands much better than previous models.<\/p>\n\n\n\n

The initial launch of the Realtime API<\/strong> offered six voice options, with two new voices added later. Now, two more new voices named Marin<\/strong> and Cedar<\/strong> have been announced. Additionally, the eight existing voices have been updated to provide a more natural and fluid experience.<\/p>\n\n\n\n

The new model also performs impressively in performance tests. In the Big Bench Audio<\/strong> test, gpt-realtime<\/strong> achieved an 82.8% accuracy<\/strong>, surpassing the 65.6%<\/strong> score of the previous model from December 2024. In the MultiChallenge Audio Benchmark<\/strong> test, the model scored 30.5%<\/strong>, exceeding the previous score of 20.6%<\/strong>.<\/p>\n\n\n\n

Along with the new model and voices, there have also been updates to the Realtime API<\/strong>. The API now supports remote MCP servers<\/strong>, visual input<\/strong>, and phone calls via Session Initiation Protocol (SIP)<\/strong>. Developers can also save and reuse their prompts.<\/p>\n\n\n\n

Despite all these improvements, OpenAI<\/strong> has lowered the price of the Realtime API<\/strong>. gpt-realtime<\/strong> is now 20% cheaper<\/strong> compared to the previous gpt-4o-realtime-preview<\/strong>. The cost is $32<\/strong> per 1M voice input tokens<\/strong> and $64<\/strong> per 1M voice output tokens<\/strong>.<\/p>\n\n\n\n

You Might Also Like;<\/h3>\n\n\n
    <\/ul>","protected":false},"excerpt":{"rendered":"

    OpenAI has introduced gpt-realtime, a new conversational model that better follows complex instructions, produces more natural-sounding voices, and is 20% cheaper. The model comes with two new voices. In October 2024, OpenAI introduced the Realtime API, which allows developers to create low-latency, multimodal experiences. Since then, thousands of developers have used this API to build …<\/p>\n","protected":false},"author":1,"featured_media":17355,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"googlesitekit_rrm_CAown96uCw:productID":"","footnotes":""},"categories":[332],"tags":[335,65],"class_list":["post-27896","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-information","tag-ai-news","tag-chatgpt-news-and-content"],"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/27896","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=27896"}],"version-history":[{"count":0,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/27896\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media\/17355"}],"wp:attachment":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media?parent=27896"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/categories?post=27896"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/tags?post=27896"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}