OpenAI cuts price of GPT-5.6 Luna, GPT-5.6 Terra

Published July 30, 2026

OpenAI said it is cutting the price of GPT-5.6 Luna by about 80% and the price of GPT-5.6 Terra by 20% after improving performance of the underlying systems.

The moves, outlined in a blog post, are notable for the following reasons:

  • Frontier AI companies are getting the message that the idea that enterprises will endure inflated token costs is bunk.
  • Open models from China as well as the US are going to apply pricing pressure to frontier labs.
  • Frontier labs can use infrastructure as a moat assuming they can continually optimize.

OpenAI said:

"Our efficiency edge comes from improving the models, the inference systems that run them, and the agentic harness that connects them to tools and context. GPT‑5.6 models take a more direct path through work. Better routing keeps hardware productive, optimized production software generates tokens more efficiently, and smarter context management helps agents avoid repeating completed work. Together, these improvements let us complete more useful work with the same compute, reducing the time, tokens, and cost required for each result."

The company added that it is using GPT-5.6 Sol is helping it deliver efficiency gains going forward. OpenAI noted that Sol "autonomously rewrote and optimized production kernels, designed and ran hundreds of experiments to improve token generation, and monitored training, intervening when problems arose."

OpenAI said API pricing will go like this:

  • $2 per million input tokens and $12 per million output tokens for Terra.
  • 20 cents per million input tokens and $1.20 per million output tokens for Luna.
  • Sol pricing remains unchanged.
  • ChatGPT and Codex subscription prices and quota budgets are unchanged.