OpenAI released its latest AI model GPT-4.5 on Friday, which will be first available to developers and ChatGPT Pro users. According to the start-up, GPT-4.5 (codenamed ‘Orion’ internally) has an improved ability to "follow user intent," and is less prone to hallucinations. OpenAI also claimed that GPT-4.5 is trained using more computing power and data than any of the AI start-up’s previous releases.
GPT‑4.5 is an example of scaling unsupervised learning by scaling up compute and data, along with architecture and optimisation innovations. This learning technique resulted in a model with broader knowledge and a deeper understanding of the world, leading to reduced hallucinations and more reliability across a wide range of topics.
Talking about the model, OpenAI chief Sam Altman said, “This isn’t a reasoning model and won’t crush benchmarks. It’s a different kind of intelligence and there’s a magic to it I haven’t felt before. I'm really excited for people to try it!”
The co-founder had previously hinted that GPT-4.5 would be the start-up's last "non-chain-of-thought" model, after which it will begin to "unify" its GPT and o-series of "reasoning" models.
The model was only trained with pre-training, supervised finetuning, and Reinforcement Learning from Human Feedback (RLHF), so this is not a reasoning model. It is an extension of the GPT series of models, unlike the ‘o’ series of models. It was trained on Microsoft Azure AI supercomputers.
The launch of GPT-4.5 comes a few days after Anthropic released Claude 3.7 Sonnet, its latest large language model and Elon Musk-led xAI unveiled Grok 3.
New Releases: Claude 3.7 Sonnet & Grok 3
A few days back, Anthropic released Claude 3.7 Sonnet, its latest AI model, and Claude Code, an agentic coding tool available in a limited research preview. The AI start-up in its blog post claimed that Claude 3.7 Sonnet is “the first hybrid reasoning model on the market” and allows users to choose between near-instant responses and extended, step-by-step reasoning.
Two weeks before that, Elon Musk's xAI launched an AI chatbot Grok 3,. It was initially available to X Premium Plus subscribers but was later made available to all users for free for a limited time. Grok 3 is an AI model that uses reasoning and pre-trained knowledge to answer questions. It is designed to be more capable than its predecessors and to improve problem-solving.
