OpenAI announced its latest model family on Thursday, introducing a powerful new set of programs to the increasingly crowded field of AI products.
GPT-5.6 comes in three versions: Sol (considered the flagship), Terra (a more intermediate option), and Luna (a budget-friendly option). These models expand what users can do in a variety of fields and promise powerful capabilities in corporate work, coding, and even scientific research.
CEO Sam Altman promises that his company’s latest model is orders of magnitude more efficient and cost-effective than previous versions, recently telling CNBC that Sol is 54% more token efficient when it comes to AI coding tasks.
Most notably, the company calls 5.6 “the most powerful cybersecurity model to date that achieves frontier performance with significantly fewer tokens.”
Indeed, there has been a lot of noise about the model’s cyber capabilities, with the Trump administration previously seeking to limit its deployment, ostensibly due to fears that it could be misused. GPT-5.6 supports defensive activities such as threat modeling, code review and patching, and blue teaming (simulating attacks on your own systems to find weaknesses before a real hacker does).
OpenAI also released a new tool called ChatGPT Work. As the name suggests, this tool is designed as a workplace companion for corporate teams, running on desktop, web, and mobile to help with everyday office tasks such as creating documents, spreadsheets, and presentations.
OpenAI’s newly announced family of models follows similar releases from competitors SpaceXAI and Meta this week.
However, GPT-5.6 and its accompanying marketing appear to be designed to target OpenAI’s primary adversary, Anthropic. Anthropic has managed to become a likable underdog in the AI race by focusing on enterprise customers and gaining share of support as a result.
Not to be outdone, OpenAI cites a notable benchmark metric, the Artificial Analytics Coding Agent Index, and claims that its latest family of models outperforms Anthropic’s models in every way.
OpenAI calls Sol “the best coding model ever” and explicitly compares it to Anthropic’s recently released (and much-hyped) Fable. Using the Coding Agent Index, OpenAI claims that Sol “sets a new bar at 80, 2.8 points above Fable 5, while using less than half the output tokens, taking less than half the time, and costing about a third less.”
Additionally, “the benefits extend across the family, with Terra slightly outperforming Fable 5 and Luna outperforming Opus 4.8.”
The company says 5.6 is now available for ChatGPT, Codex, and OpenAI APIs. The availability price per million tokens will be: Sol is $5 input / $30 output, Terra is $2.50 input / $15 output, and Luna is $1 input / $6 output.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
