OpenAI has launched the GPT-5.6 family of models for general availability, featuring Sol as its new flagship for frontier intelligence and efficiency, Terra as a balanced model for everyday work, and Luna as the most cost-efficient option. These models set new standards across various domains including coding, knowledge work, cybersecurity, and science, outperforming previous and competing frontier models with fewer tokens and lower estimated costs.

GPT-5.6 Sol demonstrates significant performance gains and improved cost-efficiency in evaluations like Agents' Last Exam and the Artificial Analysis Intelligence Index. A new 'ultra' setting is introduced for demanding tasks, coordinating multiple agents across parallel workstreams for faster completion.

Safety and security have been significantly enhanced, with GPT-5.6 featuring OpenAI's most robust safeguards to date, developed through extensive human red teaming and large-scale automated testing. In cybersecurity, the models show strong capabilities in finding and fixing vulnerabilities, aiding defenders.

GPT-5.6 is available across ChatGPT, Codex, and the OpenAI API. Developers can leverage features like Programmatic Tool Calling and a multi-agent beta. Pricing is set per 1M tokens: Sol at $5 input / $30 output, Terra at $2.50 input / $15 output, and Luna at $1 input / $6 output.

👉 Read the full article on OpenAI

OpenAI has unveiled its latest family of language models, GPT-5.6, marking a significant advancement in AI technology. This release is characterized by enhanced reasoning capabilities, improved efficiency, and superior performance across a wide range of applications.

The GPT-5.6 family includes the flagship model Sol, the balanced Terra, and the cost-efficient Luna, all designed to enable developers and enterprises to tackle more complex tasks effectively. Notable improvements are anticipated in areas such as coding, data analysis, and content generation.

This announcement reinforces OpenAI's leadership in the AI industry and is expected to further expand the capabilities of generative AI. The new models emphasize delivering better results with fewer resources, thereby promoting broader AI adoption and accessibility.

👉 Read the full article on TechCrunch

OpenAI has declared its new GPT-5.6 model as the 'preferred model' for Microsoft Copilot. This announcement comes amidst ongoing speculation regarding the partnership between OpenAI and Microsoft, signaling a continued and deepening collaboration between the two tech giants.

The integration of GPT-5.6 is expected to significantly enhance user productivity within Microsoft Copilot by leveraging the model's advanced intelligence and efficiency. This will provide more sophisticated assistance and automation across the entire Microsoft 365 suite of applications.

This strategic move underscores Microsoft's commitment to deeply embedding AI capabilities into its product offerings and highlights the central role OpenAI's cutting-edge technology plays in enterprise solutions.

👉 Read the full article on TechCrunch