OpenAI has launched GPT-6 Sol and GPT-6 Luna, two new API-only models positioned below GPT-6 Astra. Both are available immediately for developers and select ChatGPT tiers, focused on coding, professional, and high-volume everyday tasks. Sol and Luna introduce substantially lower pricing, with token costs cut about 50% versus GPT-5.6, including a larger reduction for Luna’s outputs. Benchmarks across AutomationBench, DeepSWE, FrontierCode, and OSWorld show performance competitive with Claude Opus and Fable at far lower cost. OpenAI also upgraded prompt caching, offering up to 90% discounts on cached input and new monitoring and control tools, potentially reducing infrastructure expenses and latency for long-running agents as adoption expands across products and workflows.
This update represents a notable development in the Ai sector. Organizations and founders tracking this space should evaluate potential strategic and technical implications on their operations.