OpenAI Cuts GPT-5.6 Pricing and Introduces Fast Mode
모델 출시 | Fri Jul 31 2026 00:00:00 GMT+0000 (Coordinated Universal Time) | 2 sources
OpenAI significantly reduced pricing for GPT-5.6 Luna and Terra while unveiling a Fast mode option for Sol.
Analysis
[OpenAI] cut GPT-5.6 Luna and Terra pricing [1][2]
- 80% price reduction for Luna
- 20% price reduction for Terra
- Also reflected in Codex and ChatGPT Work subscription usage
[OpenAI] introduced Fast mode in the API [1][2]
- Replaces existing Priority Processing
- Up to 2.5x faster on GPT-5.6 Sol
- Priced at 2x standard with identical intelligence level
- Automatic switching for requests tagged as priority
[GPT-5.6 Luna] expanded the cost-performance frontier [1][2]
- Matches frontier-class models from a year ago
- Reduces cost per task to approximately 6%
- Approximately 9x faster speed
- Outperforms Fable 5 on Agents' Last Exam at about 99% lower cost
[OpenAI] presented an outcome-driven intelligence matching workflow [1][2]
- Optimizes the balance of intelligence
- speed
- and cost at each step
- Coding example planning with Sol and executing with Luna
- Recommends evaluation-based model selection