OpenAI releases GPT-6 Sol and Luna models with 50% lower API prices

September 22, 2026 at 11:02 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

OpenAI released GPT-6 Sol and GPT-6 Luna, two new models now available to ChatGPT Plus, Pro, Business, Enterprise, and Edu users. The models are 50% cheaper than their GPT-5.6 predecessors and are rolling out to Codex and ChatGPT Work. GPT-6 Sol makes about half as many mistakes as GPT-5.6, while GPT-6 Luna matches GPT-5.6 Sol at about a hundredth of its cost.

Key Facts

GPT-6 Sol and GPT-6 Luna are 50% cheaper than their GPT-5.6 predecessors.
GPT-6 Sol makes about half as many mistakes as GPT-5.6.
GPT-6 Luna matches GPT-5.6 Sol at about a hundredth of its cost.
The models are available to ChatGPT Plus, Pro, Business, Enterprise, and Edu users, with Luna also available to Free and Go users in the desktop app.
OpenAI is dropping the medium GPT-6 Terra model, leaving Terra (small), Sol (large), and Astra (extra large).

Model Release and Availability

OpenAI released GPT-6 Sol and GPT-6 Luna, expanding the GPT-6 family beyond the initial GPT-6 Astra model. The models are rolling out to ChatGPT Plus, Pro, Business, Enterprise, and Edu users, with Luna also available to Free and Go users in the desktop app. Enterprise administrators must enable the new models, and availability depends on rollout and workspace settings. In ChatGPT, the models are available in Work and Codex, but not Chat.

Performance and Cost Improvements

GPT-6 Sol makes about half as many mistakes as its predecessor, GPT-5.6. GPT-6 Luna matches GPT-5.6 Sol at about a hundredth of its cost at higher effort levels. OpenAI reduced API prices for Sol and Luna by 50% compared with their GPT-5.6 promotional pricing. The company attributes the lower costs to improvements in caching and inference.

Model Positioning

OpenAI positions GPT-6 Sol for complex coding and agentic workflows, and GPT-6 Luna for focused, high-volume tasks. The GPT-6 lineup now includes Terra (small), Sol (large), and Astra (extra large), with no medium model. OpenAI trained GPT-6 Sol and Luna with similar methods as GPT-6 Astra, aiming to bring Astra's advances to faster, more affordable models.