OpenAI launches two new GPT‑6 models that offer higher cost‑efficiency and improved alignment for professional and consumer use.

OpenAI today unveiled GPT‑6 Sol and GPT‑6 Luna, two new models designed to deliver higher cost‑efficiency and stronger alignment for both professional and consumer applications.

Key Improvements in GPT‑6 Sol

GPT‑6 Sol targets enterprise workloads, offering faster inference times and reduced token pricing while maintaining the robust language understanding of its predecessor.

The model incorporates refined alignment techniques that better adhere to user intent, minimizing unintended outputs and enhancing safety for high‑stakes environments.

Consumer‑Focused Features of GPT‑6 Luna

GPT‑6 Luna is optimized for everyday users, delivering smoother conversational experiences at a lower cost per interaction.

Luna’s alignment updates focus on respectful and context‑aware dialogue, aiming to reduce bias and improve relevance in casual chats.

Performance and Pricing

  • Up to 30% lower per‑token cost compared with GPT‑5
  • Reduced latency for both models, especially on standard hardware
  • Enhanced fine‑tuning options for developers

Availability

Both GPT‑6 Sol and Luna are now accessible via the OpenAI API, with early‑access tiers for existing customers and a public rollout planned for the coming weeks.

These models represent a significant step forward in making advanced AI both affordable and responsibly aligned.

For the full announcement, see OpenAI’s Introducing GPT‑6 Sol and Luna.