Chat model
Published Updated
What is Chat model?
An AI model optimized for rapid, natural back-and-forth dialogue and immediate task execution, without an extended internal reasoning phase.
Chat models (like a standard GPT model or Claude Sonnet) are fine-tuned to act as responsive conversational agents. They excel at immediately predicting the next token, making them fast and highly effective for general drafting, basic coding, and interactive tool orchestration where low latency is required.
How it relates to Orbit
Orbit uses an included model for Apollo chat unless you configure a BYOK model in Preferences. See Complexity Tiers & Routing for details.
When you would use it
You use a chat model when you need fast, fluid interactions, general text generation, or responsive API tool-calling where a long "thinking" delay is unacceptable.
Common operations
- Powering customer service chatbots.
- Driving real-time IDE autocomplete and fast inline code generation.