Reading up on GPT-6 Luna
6 deep · digging since sep 23
- Introducing Our Smallest, Most Efficient Model: Muse Ember
Meta released Muse Ember, its smallest and most efficient model, beating Claude Haiku 5.5 and GPT-6 Luna on intelligence-per-cost on agentic coding and reasoning tasks at roughly $0.14 per task.
- Are Harnesses Worth the Cost?
A behavioral harness offers no reliability gain for frontier models like GPT-6 Sol but significantly improves consistency and cost-effectiveness for the cheaper GPT-6 Luna model.
- Building an agent harness from scratch to pass Ian Goodfellow's intelligence test - NotesByLex.com
The article walks through building a minimal AI agent harness in Python, covering model wrapping, tool registration, context compaction, and safety classification.
- Decisions API is now available in Public Beta - Announcements - OpenAI Developer Community
OpenAI’s Decisions API, now in public beta, lets apps quickly choose models, tools, or actions using text and image inputs, delivering up to 10× faster decisions via GPT‑6 Luna.
- Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war
Anthropic’s Claude Opus 5.5 and OpenAI’s GPT‑6 Sol/Luna launch with steep price cuts, undercutting prior models and triggering a new AI model price war.