OpenAI cuts GPT-5.6 Luna 80% to $0.20 per million input tokens
OpenAI dropped GPT-5.6 Luna to $0.20 per million input tokens and $1.20 per million output effective July 30, with Terra down 20% to $2 and $12 and Sol unchanged. A task that cost a dollar on last year's leading models now runs about 6 cents at nearly nine times the speed, putting high-volume production workloads in reach that were priced out before. OpenAI credits GPT-5.6 Sol optimizing its own GPU software for a 20% cut in deployment cost, though the timing tracks with pressure from low-cost Chinese providers and Microsoft pitching its MAI models as the cheaper alternative.
Source: the-decoder.com ↗
A task that cost a dollar with those models now runs about 6 cents on Luna, nearly nine times faster.
OpenAI
Why this matters
- → Production AI workloads drop from $1 to $0.06, making high-volume inference economically viable
- → Sol self-optimized its own GPU software—rare case of frontier model improving its own deployment
- → Accelerating price war with Chinese vendors and Microsoft threatens margins at labs funding trillion-dollar in
Price war reaches the core