456
Dictionary

Cost Per Token

The unit economics of LLM APIs.

1 min readupdated 2026-07-04

/ quick answer

Providers price separately for input and output tokens, usually per million. Output is typically 3-5x more expensive than input, so long context is cheap but long answers are not. The unit economics of LLM APIs.

The unit economics of LLM APIs. Providers price separately for input and output tokens, usually per million. Output is typically 3-5x more expensive than input, so long context is cheap but long answers are not. In practice: GPT-4o mini: $0.15/M input, $0.60/M output — a $0.001 support answer at 2k in + 500 out. This dictionary node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Definition
Providers price separately for input and output tokens, usually per million. Output is typically 3-5x more expensive than input, so long context is cheap but long answers are not.
Example
GPT-4o mini: $0.15/M input, $0.60/M output — a $0.001 support answer at 2k in + 500 out.
/ frequently asked

What is Cost Per Token?

Providers price separately for input and output tokens, usually per million. Output is typically 3-5x more expensive than input, so long context is cheap but long answers are not.

What is an example of Cost Per Token?

GPT-4o mini: $0.15/M input, $0.60/M output — a $0.001 support answer at 2k in + 500 out.

Why does Cost Per Token matter for AI and automation?

The unit economics of LLM APIs. It connects to the workflows, prompts and tool stacks linked on this page, so you can move from definition to execution without leaving Onexial.