OpenAI CEO Sam Altman said the company's new GPT-5.6 Sol model is 54% [1] more token-efficient on agentic coding tasks.
This improvement suggests a significant reduction in the computational cost and time required for AI to perform complex programming tasks. As enterprises integrate AI agents into their software development pipelines, efficiency gains directly impact the bottom line by lowering API costs and increasing speed.
Altman discussed the performance of the model during an interview with CNBC. He said the GPT-5.6 Sol model is as good as or better than competing models currently on the market [1, 2]. The focus on agentic coding refers to the model's ability to operate with a degree of autonomy to solve multi-step programming problems, rather than simply completing a single line of code.
Token efficiency is a critical metric for large language models because tokens represent the basic units of text processed by the AI. A 54% [1] increase in efficiency means the model can achieve the same or better results while using fewer resources. This optimization is aimed at positioning OpenAI as the preferred provider for high-scale enterprise AI deployment [3, 4].
The announcement comes as competition intensifies between OpenAI and rivals like Anthropic. By emphasizing efficiency alongside performance, OpenAI is targeting the operational overhead that often hinders the wide-scale adoption of autonomous AI agents in professional environments [3].
Altman said he did not provide specific benchmarks for other general tasks during the broadcast, focusing instead on the coding capabilities of the Sol model [1, 2].
“The new GPT-5.6 Sol model is 54% more token-efficient on agentic coding tasks.”
The shift toward 'agentic' efficiency indicates that OpenAI is moving beyond simple chat interfaces toward autonomous systems that can execute complex workflows. By reducing token consumption by more than half for coding tasks, OpenAI is attempting to solve the cost-prohibitive nature of autonomous AI agents, making it more viable for companies to replace or augment human developers with AI-driven agents.
