Advancing the Price-Performance Frontier with GPT-5.6
Source: https://openai.com/index/advancing-the-price-performance-frontier-with-gpt-5-6
Author: OpenAI
Published: 2026-07-30
Overview
OpenAI announced significant price reductions and performance improvements across the GPT-5.6 model family, claiming to advance the price-performance frontier for enterprise AI workloads.
Key Announcements
Pricing Changes (effective July 30)
- GPT-5.6 Luna (fastest, most affordable): 80% price reduction
- API: $0.20/M input tokens, $1.20/M output tokens
- GPT-5.6 Terra (balanced for everyday work): 20% price reduction
- API: $2/M input tokens, $12/M output tokens
- GPT-5.6 Sol pricing unchanged
Fast Mode (replaces Priority Processing)
- GPT-5.6 Sol Fast mode: up to 2.5x faster than Standard at 2x price
- No change in intelligence
- Backward compatible:
prioritytagged requests auto-route to Fast mode
Efficiency Strategy
OpenAI attributes gains to three layers:
1. Model efficiency — GPT-5.6 takes a more direct path through work; better routing keeps hardware productive 2. Inference systems — optimized production kernels generate tokens more efficiently 3. Agentic harness — smarter context management avoids repeating completed work
Notably, GPT-5.6 Sol itself contributed to these gains: within a human-led process, Sol autonomously rewrote production kernels (reducing serving cost by 20%) and ran experiments that increased token-generation efficiency by >15%.
Enterprise Adoption Signals
- Replit: Luna described as "intelligence too cheap to meter"
- Notion: Terra delivers comparable quality to GPT-5.5 at half cost, 60% less time
- Ramp: Luna is default model for background agent automations
- Blitzy: Luna handles 2.2x more context with 8.5x fewer output tokens at 87% lower cost
- Cognition: Luna incorporated into Devin Fusion for cost savings
- Dust: Luna is 40% faster and 40% cheaper than previous default
Model Tiers
| Model | Position | Pricing | |-------|----------|---------| | Sol | Frontier intelligence | Unchanged | | Terra | Balanced everyday work | $2/$12 per M tokens | | Luna | Fastest, most affordable | $0.20/$1.20 per M tokens |
中文概要
OpenAI 发布 GPT-5.6 系列重大价格调整：Luna 降价 80%（$0.20/$1.20 每百万 token），Terra 降价 20%（$2/$12），Sol 价格不变。同时推出 Fast mode 替代 Priority Processing，Sol Fast mode 速度提升 2.5 倍。效率改进来自三层：模型本身、推理系统、Agent 架构。值得注意的是，GPT-5.6 Sol 本身参与了优化过程，自主重写生产内核并跑实验，将服务成本降低 20%、token 生成效率提升超 15%。