New Developments

Recently, OpenAI released two new models, GPT 5.4 Thinking and GPT 5.4 Pro. GPT 5.4 Thinking is specifically designed for enterprise work, such as coding and overseeing AI agents. The new models are described as OpenAI’s most factual yet, with responses from GPT 5.4 being 18% less likely to contain errors compared to the previous version, GPT 5.2.

Moreover, individual claims made by GPT 5.4 are 33% less likely to be false compared to GPT 5.2. This improvement in accuracy is crucial for professional applications where reliability is essential.

Key Features

GPT 5.4 supports agentic activity, which means it can operate more efficiently while using less computing power. This efficiency translates into lower costs for users. OpenAI stated that GPT 5.4 can more efficiently support agentic activity, meaning it uses less computing power and therefore costs less money.

The Thinking version of GPT 5.4 introduces a new way of interacting with the model by showing an upfront plan of its reasoning. This allows users to adjust the direction of the conversation mid-response, enhancing the overall user experience.

Performance Metrics

In terms of performance, GPT 5.4 supports up to 1 million tokens of context, significantly expanding its capabilities. On GDPval, GPT 5.4 scored 83% wins or ties, while it reached 75% on OSWorld-Verified, showcasing its competitive edge in various benchmarks.

Future Implications

OpenAI positions GPT 5.4 as a step forward in making AI more reliable for professional work. Observers note that this release is a significant boost for OpenAI, potentially enhancing its standing in the competitive AI market.

Details remain unconfirmed regarding the broader implications of these advancements on OpenAI’s market position and future developments in AI technology.

The post Chatgpt 5.4 appeared first on yesdaidanews.com.