GPT-6 Sol and Luna Release — Drastic Cost Reduction & Performance Optimization

Release of GPT-6 Sol и Luna

The launch of new specialized models represents an aggressive move towards resource optimization으로 lowering costs while maintaining advanced technical performance.

Key Arguments

  • Extreme Cost Efficiency/Price Drop: The price (Sol) has dropped to $2/$10 per million tokens; however, even more significantly, certain versions (Luna compared to original 5.6) have seen prices slashed by up to 10x ($0.1/$0.5 per million tokens).
  • Performance Parity with Lower Resources: Even at maximum settings, Luna produces results comparable enoughsto Sol running at minimum power levels but consumes up to 10 times fewer resources. Additionally, Sol can solve specific tasks nearly 5 times cheaper than Opus 5.
  • Functional Improvements: These models inherit valuable features from Astra—specifically improved programming skills, better computer system interaction, and enhanced non-trivial problem-solving abilities.
  • User Incentives: Subscribers for Claude or ChatGPT receive one free limit reset (

! DYOR (Do Your Own Research)