OpenAI is lowering prices and increasing performance for its GPT-5.6 models. GPT-5.6 Luna will see an 80% price reduction, making it significantly more cost-effective for high-volume tasks. GPT-5.6 Terra will be 20% cheaper, offering a balanced option for everyday use. These price adjustments apply to API usage, as well as Codex and Chat GPT Work.

In addition to cost savings, GPT-5.6 Sol will feature a new "Fast mode" in the API, delivering up to 2.5x faster speeds at twice the price of standard processing. This allows customers to optimize for speed when needed. These efficiency gains stem from improvements in model architecture, inference systems, and agentic harnesses, enabling more work to be done with the same compute resources.

The GPT-5.6 family now offers greater flexibility, allowing businesses to match intelligence, speed, and cost to specific outcomes. This move aims to make advanced AI more accessible and affordable, aligning with OpenAI's mission to ensure AGI benefits all of humanity. The new pricing and Fast mode will begin rolling out on July 30th.