GPT-5.6 Luna and Terra are now available

Two new OpenAI models are now available to power your chatbot: GPT-5.6 Luna and GPT-5.6 Terra.

GPT-5.6 Luna is our new default, and it's on every plan. It's a fast-tier model, so answers stay quick, but it's noticeably better at one thing that matters for support: knowing when it can't help. When a question falls outside your docs, Luna reliably offers a human handoff instead of guessing or going quiet. Earlier fast models often did neither.

GPT-5.6 Terra is the advanced-tier option, on Business plans and up. It's the latest generation for deep, technical answers over dense references, when you want the strongest possible response and not just the fastest.

To switch, open Settings > Chatbot > Model and choose GPT-5.6 Luna or GPT-5.6 Terra.

The models docs list every option and what each is good at.

What this means for docs teams

A documentation chatbot should refuse and hand off when the docs do not contain the answer, and fast-tier models are now good enough to do that reliably. If ticket deflection with a graceful human fallback is what you care about, Luna is the safe default on any plan. Save Terra for dense API references where answer depth, not speed, is the bottleneck.