OpenAI rolled out a research preview of GPT-4.5, it's 'largest and most knowledgeable model yet,' on Thursday.
Just two weeks after CEO Sam Altman shared the roadmap for OpenAI's upcoming AI models, the company has officially launched GPT-4.5. In a new system card, OpenAI says that GPT-4.5 is its 'largest and most knowledgeable model yet.' Early testing of the model indicates that interacting with GPT-4.5 is more natural and that it is better suited for writing, programming, and solving practical problems than GPT-4o. Plus, the new model is less prone to hallucinations than its predecessors.
' It has better world knowledge, improved writing abilities, and a more refined personality, but no 'net-new frontier capabilities compared to previous reasoning releases.' The company also says that despite its improvements in some areas, 'its performance is below that of o1, o3-mini, and deep research on most preparedness evaluations.' As of this Thursday, anyone with a ChatGPT Pro subscription can select GPT‑4.5 in the model picker on the web, mobile, and desktop. GPT‑4.
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
OpenAI Unveils GPT-4.5, Available To ChatGPT Pro UsersChatGPT is getting even smarter. OpenAI has unveiled its latest model, GPT-4.5. The new model is already available to ChatGPT Pro users and developers.
Read more »
OpenAI Charts Course for Future with GPT-4.5 and GPT-5OpenAI CEO Sam Altman outlines the company's roadmap for upcoming AI model releases, prioritizing unified intelligence and user-friendly experiences.
Read more »
OpenAI Accuses Chinese Rival DeepSeek of Data Theft for AI Model TrainingOpenAI, the creator of ChatGPT, alleges that Chinese AI company DeepSeek used OpenAI's data to train its competing models, potentially violating OpenAI's terms of service. Microsoft security researchers detected suspicious data exfiltration from OpenAI accounts linked to DeepSeek. OpenAI claims evidence of 'distillation,' a technique where smaller models are trained using data from larger ones, suggesting DeepSeek leveraged OpenAI's expensive GPT-4 training data. While OpenAI acknowledges its own past use of web data without explicit consent, it emphasizes the need to protect its intellectual property and calls for collaboration with the US government to safeguard advanced AI technology.
Read more »
OpenAI's GPT-5 to Integrate o3 Reasoning, Offer Free AccessOpenAI CEO Sam Altman announced in a roadmap that the upcoming GPT-5 will integrate the o3 reasoning model and be accessible to free users. GPT-4.5, the next iteration, will be the last non-chain-thought model before GPT-5. OpenAI aims to unify its o-series and GPT-series models, creating systems that can utilize all its tools and handle complex tasks. GPT-5 will integrate o3 and other advanced technologies, with free users having standard access and paid tiers offering higher intelligence levels.
Read more »
OpenAI's GPT-4: A Leap Forward in AIOpenAI Unveils GPT-4, Its Advanced AI Model with Enhanced Capabilities, Revolutionizing Industries and Sparking Ethical Debate
Read more »
OpenAI Simplifies AI Offerings, Phases Out o3 in Favor of Unified GPT-5OpenAI CEO Sam Altman announced a significant shift in the company's AI strategy, moving away from the complex o3 model and towards a simplified, unified GPT-5. This decision aims to streamline user experience and provide a more intuitive AI interaction. GPT-4.5, codenamed Orion, will serve as the final non-chain-of-thought model before GPT-5's release, which will integrate various technologies, including o3. OpenAI aims to create a system that seamlessly combines o-series and GPT-series models, offering a wider range of capabilities across voice, canvas, search, and deep research. The move comes as competition intensifies with DeepSeek, a Chinese startup whose R1 reasoning model rivals o3's performance at a lower cost.
Read more »