GPT-5.6 received the green light and arrives with a key novelty: it's not a single model, but three different profiles —Sol, Terra, and Luna— designed for different needs. If you already use ChatGPT at work, understanding which one is best for each case can change your productivity. Here we explain what each one is and how to take advantage of them.

The announcement, reported on July 7 by media like Axios and The Verge, comes after the U.S. government's review and closes weeks of uncertainty for the industry. Beyond the noise, what's relevant for professionals in the region is practical: now you can choose the model according to the task, balancing quality, speed, and cost.

Sol, Terra, and Luna: three profiles, one objective

The logic behind the GPT-5.6 family is to stop using a single model for everything. Each profile is optimized for a scenario:

• Sol (flagship): the most capable, designed for complex reasoning tasks, deep analysis, and demanding creative work. It's the one you choose when quality trumps cost.

• Terra (volume): the balanced one, designed to process large volumes of tasks with good quality and controlled cost. Ideal for workflows and automations.

• Luna (fast and economical): the agile one, optimized for fast and low-cost responses. Perfect for simple tasks, frequent queries, and prototyping.

How to choose the model according to the task

What it means for professionals who already use ChatGPT

The change of mindset is key: instead of using the most powerful model for everything, it's a good idea to assign the right model to each task. Writing a quick email doesn't need the same engine as designing a strategy. This logic of "the right model for each job" reduces costs in intensive uses and improves speed. For teams that automate processes, the difference shows in the bill and in the response times. If you work with automated flows, you'll be interested in understanding how they combine with AI agents by industry.

To follow the detail of the announcement and its implications, media like The Verge and Axios have been covering the evolution of OpenAI's models. And to gauge the context of enterprise adoption, McKinsey's State of AI shows how much the uses of generative AI grew in organizations.

Recommended Coderhouse courses

Taking good advantage of a model family like GPT-5.6 requires knowing when and how to apply it. At Coderhouse you can train according to your profile:

• Introduction to Artificial Intelligence Course: to understand the fundamentals and use ChatGPT with judgment.

• AI Engineering Course: if you want to build solutions that leverage different models according to the task.

• AI Agents Course: to create agents that orchestrate fast and powerful models intelligently.

Get the most out of AI: train to choose the right model for each task and turn ChatGPT into a true productivity multiplier.

Frequently asked questions

What is GPT-5.6?

It's OpenAI's new model family, made up of three profiles —Sol, Terra, and Luna— optimized for different needs of quality, volume, and speed, instead of a single model for everything.

How do Sol, Terra, and Luna differ?

Sol is the most capable for complex tasks; Terra balances quality and cost for volume and automations; Luna is the fastest and most economical for simple tasks and frequent queries.

Which one is best to use at work?

It depends on the task: Sol for analysis and strategy, Terra for processes at scale, and Luna for quick queries. Assigning the right model to each use optimizes cost and speed.

Do I need technical knowledge to take advantage of them?

For everyday use in ChatGPT, no. To integrate them into automations or products, it's a good idea to train in AI fundamentals and applied engineering.