GPT-6 Sol and Luna: What OpenAI’s Cheaper Models Mean for AI Agents
OpenAI just cut GPT-6 Sol and Luna API prices and caching costs. Here's what changes for agent builders.
Why it matters
OpenAI released two new models (Sol and Luna) with lower API pricing and cheaper prompt caching, specifically positioned to reduce the operational cost of running AI agents. For practitioners building or deploying agentic workflows, this directly affects token economics and caching budgets — a material shift in the cost basis for production agent deployments.
The key facts
6 to knowTwo new models: GPT-6 Sol and Luna
Lower API prices (specific pricing not disclosed in article)
Cheaper caching feature (pricing structure not disclosed)
Stronger coding results claimed
Target use case: reducing cost of AI agents
Published Sep 25, 2026
The story so far
Earlier coverage of this storyline
- Introducing GPT-6 Sol and LunaOpenAI Blog
- OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakesTechCrunch AI
- Bring more intelligence to everyday work with GPT-6 Sol and GPT-6 Luna on Amazon BedrockAWS Machine Learning Blog
- OpenAI's GPT-6 Sol and Luna cut prices in half but barely move the needle on performanceThe Decoder
- Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price warSimon Willison
- This story
Go to the source
TechRepublictechrepublic.com
Publisher excerpt: OpenAI launched GPT-6 Sol and Luna with lower API prices, cheaper caching and stronger coding results aimed at reducing the cost of AI agents.