Yesterday OpenAI unveiled GPT-4.5 — its newest and most advanced language model yet. The model is positioned as the company's "most knowledgeable model" to date, offering improved capabilities and more natural interaction with users.
Key features of GPT-4.5
Improved capabilities
- More natural conversation: Interacting with GPT-4.5 feels "warmer" and more intuitive, with an improved ability to understand context and nuance.
- Broader knowledge base: The model has more extensive world knowledge compared to previous versions.
- Fewer hallucinations: GPT-4.5 generates accurate information 63% more often than previous versions.
- Improved "emotional intelligence": The model is better at recognizing emotional nuance and responding to it more appropriately.
How it differs from other OpenAI models
GPT-4.5 is not a "reasoning model" like the "o" series (o1, o3-mini). It doesn't use step-by-step chain-of-thought reasoning, but it outperforms previous models on creative tasks, writing, and practical problem-solving.
In the words of Sam Altman, CEO of OpenAI:
"This is the first model that feels to me like talking to a thoughtful person. I've had several moments where I've leaned back in my chair, amazed at getting real, good advice from an AI."
What's next for the model
OpenAI notes that GPT-4.5 may be the last model before the move to GPT-5, which will integrate various company technologies, including the o3 reasoning model.
GPT-4.5 marks a significant step forward for OpenAI's language models, offering more natural and intuitive interaction, though with some limitations in complex reasoning and logic.
Try the new model right now in GPTunneL! Our team quickly adds all the top AI models from around the world to the service. Subscribe to our Telegram channel to stay up to date with the latest updates!