• Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法

​
​よろず知財コンサルティングのブログ

OpenAIの新モデル「o3」と「o4-mini」

17/4/2025

0 Comments

 
4月16日、OpenAIの新モデル「o3」と「o4-mini」がリリースされました。
新モデルは、「考える力」と「道具を使う力」を統合し、複雑な問題解決やマルチモーダル推論を可能にしたということで、大規模強化学習による思考訓練や安全性への配慮、実用的なツール連携が特徴ということです。
o3とo4-miniはそれぞれ、o1とo3-miniの進化版ということですが、「o3とo4-miniは、これまででもっとも知的なモデル」というOpenAIの表現ですので、最近、GoogleのGemini 2.5 Proが推論能力・処理速度等々でOpenAIのモデルを抜き去っていましたが、またOpenAI優位に戻るのでしょうか?
ちなみに、最新(2025年4月17日現在)のTrackingAIのIQ (Mensa Norway) では、OpenAI o3のIQが136で、Gemini 2.5 Pro のIQが128でした。
 
April 16, 2025
Introducing OpenAI o3 and o4-mini
Our smartest and most capable models to date with full tool access
https://openai.com/index/introducing-o3-and-o4-mini/
 
 
 
OpenAI’s New Models “o3” and “o4-mini”
 
On April 16, OpenAI released its new models “o3” and “o4-mini.”
These new models are said to integrate “the power of reasoning” and “the ability to use tools,” enabling complex problem-solving and multimodal reasoning. Their key features include large-scale reinforcement learning for training reasoning abilities, safety considerations, and practical integration with tools.
o3 and o4-mini are described as evolved versions of o1 and o3-mini, respectively. According to OpenAI, “o3 and o4-mini are the most intelligent models to date.”
Recently, Google’s Gemini 2.5 Pro had surpassed OpenAI’s models in areas like reasoning performance and processing speed — but could this mean OpenAI is reclaiming its lead?
As of the latest update (April 17, 2025) from TrackingAI's IQ (Mensa Norway), OpenAI o3 has an IQ of 136, while Gemini 2.5 Pro has an IQ of 128.

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments



Leave a Reply.

    著者

    萬秀憲

    アーカイブ

    June 2025
    May 2025
    April 2025
    March 2025
    February 2025
    January 2025
    December 2024
    November 2024
    October 2024
    September 2024
    August 2024
    July 2024
    June 2024
    May 2024
    April 2024
    March 2024
    February 2024
    January 2024
    December 2023
    November 2023
    October 2023
    September 2023
    August 2023
    July 2023
    June 2023
    May 2023
    April 2023
    March 2023
    February 2023
    January 2023
    December 2022
    November 2022
    October 2022
    September 2022
    August 2022
    July 2022
    June 2022
    May 2022
    April 2022
    March 2022
    February 2022
    January 2022
    December 2021
    November 2021
    October 2021
    September 2021
    August 2021
    July 2021
    June 2021
    May 2021
    April 2021
    March 2021
    February 2021
    January 2021
    December 2020
    November 2020
    October 2020
    September 2020
    August 2020
    July 2020
    June 2020

    カテゴリー

    All

    RSS Feed

Copyright © よろず知財戦略コンサルティング All Rights Reserved.
サイトはWeeblyにより提供され、お名前.comにより管理されています
  • Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法