OpenAI Launches GPT-6 Sol and Luna Models
Both models are available via API at 50% lower prices than GPT-5.6 and are rolling out in Codex and ChatGPT Work.
- On DeepSWE v1.1, GPT-6 Sol nearly matches Claude Fable 5 at roughly 80% lower cost per task.
- GPT-6 Luna matches Fable 5 (medium) on DeepSWE v1.1 at 96% lower cost per task.
- On AutomationBench, GPT-6 Sol leads Fable 5.1 with 88% lower reported cost per task.
- Improved prompt caching keeps more context reusable across reasoning efforts and tool calls.
- Access is rolling out to Plus, Pro, Business, and Enterprise users.
Developers and enterprise users can run coding agents and multi-app business workflows at half the API cost of the previous generation.
