GPT-5.6 Luna API Pricing & Cost Calculator

How this is calculated

GPT-5.6 Luna is priced at $0.20 per million input tokens and $1.20 per million output tokens at OpenAI's published list price, with a 90% prompt caching discount ($0.02 per million cached reads). That's 10x cheaper than GPT-5.6 Terra ($2/$12) on input and 10x cheaper on output, putting it in the high-volume utility tier alongside MiniMax-M3. Note: OpenRouter currently runs a limited-time 50% discount that can show Luna at $0.10/$0.60 in live calculators.

Verdict

Luna is the floor of the GPT-5.6 lineup at $0.20/M input and $1.20/M output. The 90% cache discount takes cached input to $0.02/M, which on a stable system prompt reused across millions of daily classification or formatting requests is essentially free input cost. A representative 30K input + 1K output utility call lands at roughly $0.0072 before discounts and $0.0011 with a 90% cache hit, so the per-call cost on the high-volume utility tier falls below what most pipelines cost to manage. The 10x gap above Luna in the lineup (Terra at $2/$12) is what makes the price ceiling for routing: anything more complex than a one-shot classifier belongs upsold.