Text
The core chat-and-generation family. You send a prompt, you get a reply. Text models vary in size, speed, context window, and cost, so browse the catalog and use the decision guide.Reasoning
A capability of certain text models. They work through a problem step by step before answering, and you control how much effort they spend. The catalog marks which models support it, and the thinking counts toward your usage.Vision
Vision-capable text models accept images alongside text: describing a photo, reading a chart. Input only; the model replies in text. Accepted formats are PNG, JPEG, and WebP. Documents (PDF, DOCX, PPTX, XLSX, XLS, text) are parsed to text first and then sent as text, so any model can read them. Use vision for pictures, the file parser for documents. Audio and video are not available today, on any surface.Related
Choosing a Model
From the task to the right model.
Model List
The whole catalog.