OpenAI has broadly released its GPT-5.6 Sol, Terra, and Luna models following an initial limited launch, according to CEO Sam Altman.

This update targets the high cost and complexity of agentic coding, which allows AI to operate more autonomously. By reducing the number of tokens required to complete complex programming tasks, OpenAI aims to lower operational costs for enterprises and improve the overall safety of AI-generated code.

During an interview with CNBC, Altman said the GPT-5.6 Sol model is 54% [1] more token-efficient on agentic coding tasks compared to rival models. This efficiency allows the system to process and generate code using fewer computational resources, a critical factor for companies scaling AI integration into their software development pipelines.

Altman said the company released the three new models—Sol, Terra, and Luna—broadly on Thursday [2]. The rollout follows a period of limited availability where the models were tested in controlled environments to ensure stability and performance.

The push for token efficiency is part of a broader strategy to make large language models more sustainable. As AI agents take on more complex, multi-step coding projects, the volume of data processed can grow exponentially, leading to higher latency and increased spending for the end user.

By optimizing the Sol model specifically for agentic tasks, OpenAI is positioning its latest release as a tool for high-efficiency engineering. The company intends for these improvements to streamline how developers interact with AI, reducing the overhead associated with long-form code generation and iterative debugging.

“"GPT-5.6 Sol is 54% more token efficient on agentic coding tasks."”

Increased token efficiency represents a shift from focusing solely on raw intelligence to optimizing the cost of intelligence. For the enterprise market, a 54% reduction in token usage for coding tasks can significantly lower the monthly API spend and reduce the carbon footprint of AI operations, making autonomous coding agents more commercially viable at scale.