Reasoning Models
LLMs that spend extra compute thinking before they answer.
Reasoning models (o1, o3, DeepSeek-R1, Gemini Thinking) trade latency for accuracy on hard problems like coding, math and analysis.
You're offline. Check your connection — we'll retry automatically.