OpenAI has laid down the latest marker in the AI pricing war.
On Thursday, OpenAI CEO Sam Altman touted the company’s “major price cuts,” which came in the form of cheaper usage of GPT-5.6 Luna and Terra, two of the company’s latest AI models.
“We want to offer the best price/intelligence tradeoff at every level,” Altman wrote on X.
major price cuts today:
*80% drop for GPT-5.6 Luna, now $0.20 per million input tokens and $1.20 per million output
*20% drop for GPT-5.6 Terra, to $2/$12
*GPT-5.6 Sol gets Fast mode in the API, up to 2.5x the speed for 2x the price, same intelligence pic.twitter.com/erC6u4VoDR— Sam Altman (@sama) July 30, 2026
OpenAI said that Luna would cost 80% less to use and Terra would cost 20% less.
“These lower prices for Luna and Terra are also reflected in how usage is counted against paid subscriptions when using Codex and ChatGPT Work,” the company said in the announcement.
EMARKETER senior analyst Jacob Bourne said the announcement was another sign that “the era of tokenmaxxing is over.”
“Enterprises have figured out how easy it is to burn tokens without getting value back, and they’re pushing back on those increasing AI bills,” Bourne told Business Insider, which is a sister company to EMARKETER.
OpenAI said the improvements come from its focus on everything from the model itself to harnesses, an industry term for the software around the AI model that connects it to things like tools. The company and much of the industry have also worked on routing to ensure they use their available compute more efficiently.
“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,” the company said in a statement. “Better routing keeps hardware productive, optimized production software generates tokens more efficiently, and smarter context management helps agents avoid repeating completed work.”
GPT-5.6-Sol, OpenAI’s current frontier model, was not included in the announcements. OpenAI rolled out the 5.6 series of models roughly three weeks ago after pausing a broader rollout at the US government’s request.
Pricing remains one of the biggest topics hanging over AI. Moonshot AI’s recent release of the open-weight Kimi K3 model has increased pressure on closed-model makers like OpenAI. At the same time, other competitors like Anthropic are trying to strike a balance between subscription and usage-based pricing and a finite amount of available compute.
Altman and other AI CEOs have said they constantly hear from companies about whether they are getting a sufficient return on their AI investment.
Other companies have made similar pushes. Many in the industry expect token prices to decrease in the long run.
Google has recently touted the cost efficiency of its models. During a quarterly earnings call on Wednesday, Microsoft CEO Satya Nadella repeatedly stressed that “cost efficiency” was at the core of its MAI-Thinking-1 model.
“We are building a new model system where the harness, context, memory, and action space are separate from any one model family, thereby moving the frontier on the cost to outcome curve,” Nadella said.
Read the full article here
