OpenAI has unveiled its newest family of models, with GPT-5.6 leading the charge as a more efficient and cost-effective solution for enterprise use, coding, and scientific research.

Celebrating greater token efficiency and enhanced cybersecurity features, the new models aim to outshine their competitors, particularly Anthropic, in areas like code review and patching. CEO Sam Altman highlighted Sol’s impressive performance at 54% more token efficient during AI coding tasks, a significant improvement over previous versions.

For enterprise teams, OpenAI has introduced ChatGPT Work, a tool designed to assist with daily clerical tasks such as document drafting and presentations, running seamlessly on desktop, web, and mobile platforms. However, the focus remains on Sol, Terra, and Luna, each offering different price points: $5 input / $30 output for Sol, $2.50 input / $15 output for Terra, and $1 input / $6 output for Luna.

The launch of GPT-5.6 comes amidst heightened competition in the AI market, with SpaceXAI and Meta also releasing new models this week. OpenAI’s claim to outperform Anthropic in the Artificial Analysis Coding Agent Index adds fuel to the ongoing race, solidifying their position as a formidable force in the tech landscape.