OpenAI unveiled its latest household of fashions on Thursday, introducing a brand new set of heavyweight packages into an more and more crowded subject of AI choices.
GPT-5.6 is available in three variants: Sol (thought of its workhorse), Terra (a extra intermediate choice), and Luna (its price range pleasant choice). These fashions broaden what customers can do throughout quite a lot of fields — with the corporate promising highly effective capabilities in enterprise work, coding, and even scientific analysis.
CEO Sam Altman has promised that his firm’s latest fashions are orders of magnitude extra environment friendly and cost-effective than earlier variations, just lately telling CNBC that Sol is 54% extra token environment friendly relating to AI coding duties.
Most notably, the corporate calls 5.6 its “strongest cybersecurity mannequin but, attaining frontier efficiency with considerably fewer tokens.”
Certainly, a lot hubbub has been made in regards to the mannequin’s cyber capabilities, because the Trump administration beforehand sought to restrict its rollout, ostensibly resulting from fears of how the mannequin could possibly be misused. GPT-5.6 helps defensive actions, together with risk modeling, code assessment and patching, and blue teaming (simulating an assault by yourself methods to search out weaknesses earlier than actual hackers do).
OpenAI additionally launched a brand new instrument referred to as ChatGPT Work, which — simply because it sounds — is designed as a office companion for enterprise groups, working on desktop, net, and cell, that may assist with day by day clerical duties, like drafting paperwork, spreadsheets, and shows.
OpenAI’s newly introduced household of fashions follows on the heels of comparable releases this week from rivals SpaceXAI and Meta.
Nevertheless, GPT-5.6 and its attendant advertising and marketing appears most designed to take purpose at OpenAI’s major opponent, Anthropic. Anthropic has managed to make itself the likable underdog of the AI race, focusing fixedly on enterprise clients and successful a rising share of assist because of this.
To not be outdone, OpenAI cites the Artificial Analysis Coding Agent Index, a notable benchmarking metric, to assert that its newest household of fashions outshines Anthropic’s fashions at each flip.
OpenAI calls Sol its “finest coding mannequin but,” and has explicitly in contrast it to Anthropic’s recently released (and much hyped) Fable. Utilizing the Coding Agent Index, OpenAI claims that Sol “units a brand new cutting-edge at 80, 2.8 factors above Fable 5, whereas utilizing lower than half the output tokens, taking lower than half the time, and costing about one-third much less.”
It provides: “That benefit extends throughout the household: Terra performs simply above Fable 5, whereas Luna outperforms Opus 4.8.”
The corporate says that 5.6 is now accessible throughout ChatGPT, Codex, and the OpenAI API. Availability per million tokens is priced as follows: Sol is $5 enter / $30 output, Terra is $2.50 enter / $15 output, and Luna is $1 enter / $6 output.
If you buy by hyperlinks in our articles, we may earn a small commission. This doesn’t have an effect on our editorial independence.
