GPT models

What each GPT model is good at, which two you can use in our chat, and where the rest of OpenAI's GPT-6 family fits.

"GPT" is the name OpenAI gives to its family of large language models. The current generation, GPT-6, arrived in September 2026 and comes in several tiers, from fast and inexpensive to slow and very capable. This section explains the models you can use on GPT.me, the ones you cannot, and how they fit together.

In our chat

  • GPT-6 Luna is the default model here. OpenAI positions it as the fast, cost-efficient tier of GPT-6. It reads images and suits everyday questions, writing and translation.
  • GPT-6.1 Terra is the mid-tier model released in October 2026. In our chat it reasons briefly before answering, and the Think chip switches to it. It also reads images.

Not available in our chat

  • GPT-6 Sol is OpenAI's model for complex professional work, more capable than Luna and Terra. It is not available in our web chat, but you can use it in our AI Chat app and on ChatPlus.
  • GPT-6 Astra is OpenAI's flagship model, available through OpenAI's own products and API. It is not offered on GPT.me.

Learn more

  • What is GPT? explains what the letters mean, how these models work and who makes them.
  • GPT history follows the models from GPT-1 in 2018 to GPT-6.1.
  • GPT online compares the ways to use GPT in a browser.
  • GPT for free lists honest ways to use GPT without paying, and their limits.
  • GPT API is for developers who want to call the models from their own code.

In our chat

Learn

Comparisons

Got a question? Ask it now

No account and no download. Open the chat, type what is on your mind, and the answer starts appearing in a second or two.

Start chatting