Inference
Using an already-trained AI to get an answer.
In everyday terms
Every time you send ChatGPT a message, it runs inference. The model's knowledge is frozen. It just applies what it learned to your input.
For professionals
A forward pass through fixed weights. Cost is per request (per token for LLMs) and dominates the bill for popular products.
Think of it like…
Training is going to cooking school. Inference is cooking dinner tonight.
You've already seen it
AI pricing "per million tokens" is the price of inference.
Myth vs reality
Myth: Using AI is free once it's built.
Reality: Every answer costs computing power. That's why AI tools have usage limits.
Quick check
You ask a chatbot a question. Is that training or inference?
- Training
- Inference
- Both equally
- Neither
Show answer
Inference: Using a trained model to answer is inference.