OpenAI said on July 30 that it cut application programming interface (API) prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% after reducing the models’ operating costs.

The company also introduced a Fast mode for GPT-5.6 Sol that delivers up to 2.5x the speed of Standard processing at 2x the price.

Luna now costs US$0.20 per million input tokens and US$1.20 per million output tokens.

Terra now costs US$2 per million input tokens and US$12 per million output tokens. Sol pricing is unchanged.

OpenAI previously described GPT-5.6 as a three-model lineup, with Sol as its workhorse model, Terra as a mid-tier option, and Luna as a lower-cost model.

It also said Sol is its best coding model yet.

OpenAI said the lower Luna and Terra pricing will also reduce how many credits those models use in Codex and ChatGPT Work.

ChatGPT and Codex subscription prices and quota budgets remain unchanged.

Fast mode replaces Priority Processing in the API, and existing API requests tagged as priority will automatically use Fast mode.

🔗 Source: OpenAI

Recent OpenAI developments