Inside GPT-5.6: What OpenAI's newest AI models Sol, Terra and Luna bring to the table

OpenAI has introduced GPT-5.6 Sol, its newest flagship artificial intelligence model, alongside two additional models, Terra and Luna, as it pushes the boundaries of AI capabilities. The launch, however, comes with an unusual twist. Instead of making the model immediately available to all users, OpenAI is beginning with a limited rollout to a small group of trusted partners while working closely with the US government.The release shows two parallel trends shaping the AI industry, one is, increasingly powerful models and the other is, growing regulatory scrutiny over how they are deployed.What is GPT-5.6 Sol?GPT-5.6 Sol is OpenAI's most advanced AI model so far. According to the company, it is designed for tasks that require deeper reasoning, long-term planning and technical expertise.The model is built to perform better than previous generations in software development, scientific research, biology, genomics and cybersecurity. OpenAI also says Sol delivers stronger coding performance, setting a new benchmark on Terminal-Bench 2.1, an industry evaluation for programming capabilities.One of its biggest additions is enhanced reasoning. Sol includes a "max" reasoning mode for more detailed analysis and an "ultra" mode that uses multiple AI sub-agents to solve complex problems.Meet Terra and LunaGPT-5.6 is not a single model but a family of AI systems designed for different workloads.Sol is the flagship model built for advanced reasoning, research and technical applications.Terra is positioned as a balanced model aimed at everyday professional work, offering a mix of capability and efficiency.Luna is the fastest and most affordable model in the lineup, designed for applications where speed and lower operating costs matter.The three-tier strategy mirrors a broader industry trend of offering different AI models for different user needs instead of relying on a single flagship system.Why isn't everyone getting access immediately?Unlike previous launches, GPT-5.6 will not be available to all users from day one.OpenAI has started with a preview programme that gives access only to a limited group of trusted partners through its API and Codex platform. Broader availability for ChatGPT users and developers is expected in the coming weeks.The restricted rollout follows discussions between OpenAI and the US government over the release of increasingly capable AI systems.According to The Washington Post, the US government will vet organisations seeking early access to GPT-5.6 while a broader framework for future AI releases is developed.OpenAI has described the arrangement as temporary, saying it does not believe government-approved access should become the long-term standard for deploying frontier AI models.Why is the US government involved?The limited rollout highlights how rapidly AI regulation is evolving.As AI models become more capable of writing software, identifying cybersecurity vulnerabilities and handling complex scientific tasks, governments have become increasingly concerned about how these systems could be misused.The latest move marks a shift in the US government's approach, reflecting growing interest in balancing innovation with national security and public safety.How much does GPT-5.6 cost?OpenAI has announced pricing for all three models.GPT-5.6 Sol: $5 per million input tokens and $30 per million output tokens.Terra: $2.50 per million input tokens and $15 per million output tokens.Luna: $1 per million input tokens and $6 per million output tokens.Initially, these models will be available through OpenAI's API and Codex, with wider access expected later.Why GPT-5.6 mattersThe launch of GPT-5.6 is about more than another AI upgrade.It highlights how the next phase of artificial intelligence is being shaped by two competing forces. On one hand, AI companies are racing to build more capable systems for coding, research and enterprise applications. On the other, governments are becoming more involved in deciding how and when those systems reach users.With Sol, Terra and Luna, OpenAI is expanding its AI portfolio while also offering an early glimpse into how frontier AI releases may increasingly be accompanied by regulatory oversight rather than immediate public access.Also read: ₹13 crore salaries and counting: Anthropic's H-1B filings reveal the cost of winning the AI race