GPT-5.6 long-context cost calculator

Estimate token charges for GPT-5.6 Sol, Terra, or Luna. The calculator automatically changes to the long-context rate when the entered input total is greater than 272,000 tokens.

Important: enter the three input categories as non-overlapping billed token counts. This worksheet estimates model token charges only. It excludes tool-call charges, regional processing uplifts, contractual discounts, credits, taxes, and any later pricing change.

Rate table used by this calculator

USD per 1 million tokens, checked August 31, 2026
Model / mode	Band	Input	Cached	Cache write	Output

Formula

(uncached input × input rate + cached input × cached rate + cache writes × write rate + output × output rate) ÷ 1,000,000

The context band is selected from the total entered input: uncached input + cached input + cache-write tokens. Exactly 272,000 uses the short-context row; a total greater than 272,000 uses the long-context row.

Sources and scope

Rates come from the official OpenAI API pricing page. The threshold and context limits are documented on the official GPT-5.6 Sol, Terra, and Luna model pages. Fast configuration and effective-tier behavior are documented in the Fast mode guide.

Download the 12-row pricing matrix (CSV)