(Bloomberg) -- OpenAI is set to release its most advanced model widely on Thursday and is expanding its preview globally, the company said.
Most Read from Bloomberg
Greece Offers Bounty to Catch Ravenous Fish Lured by Warming Sea - Chip Stocks Sink After Blistering Run as Oil Jumps: Markets Wrap - Two Millennium Trading Pods Made About $3.7 Billion Last Month - China Sentences Official to Death Over $325 Million in Bribes
OpenAI initially rolled out a version of its GPT-5.6 only to select partners, following pressure from the Trump administration to stagger the release. The government has since lifted restrictions on the platform, Axios reported. This week's launch comes with a green light from US government leadership, an OpenAI spokesperson said.
The moves comes after archrival Anthropic clashed with Washington over the potential misuse of its Mythos and Fable models. Those were put under unusual export restrictions by the Department of Commerce — which were eventually lifted after extensive consultation. That action rippled across the AI industry and raised questions over how much control governments would exercise over frontier AI.
A spokesperson for the Commerce Department did not respond to requests for comment.
The GPT-5.6 family of artificial intelligence models from OpenAI spans three tiers. The flagship Sol product is the most capable, Terra is balanced for everyday work and costs half as much per token, and Luna is designed to be fast and affordable, according to the company. OpenAI's signature ChatGPT is built atop the GPT foundation, which other developers also use to create AI tools and services.
--With assistance from Maggie Eastland. Most Read from Bloomberg Businessweek
At 17, She Sued Meta and Google, and Won. Now She's Ready to Tell Her Story - 'Southern Squeeze' Grips US Cities Once Known for Affordability - Dan Carlin Believes That American History Rhymes, Not Repeats
©2026 Bloomberg L.P.