OpenAI Slashes API Prices for GPT-5.6 Lineup As Efficiency Gains Pay Off

• API pricing for the lightweight GPT-5.6 Luna model drops by 80%, while the middle-tier GPT-5.6 Terra sees a 20% price reduction. 
• A new API "Fast mode" for the flagship GPT-5.6 Sol delivers up to 2.5 times higher speeds at double the standard rate, replacing the previous Priority Processing tier. 
• Efficiency breakthroughs were partly driven by GPT-5.6 Sol itself, which autonomously rewrote production code kernels to cut model serving costs by 20%. 

OpenAI announced significant cost reductions and performance upgrades across its GPT-5.6 artificial intelligence model family on Thursday, citing efficiency gains that led to lower expenses for enterprise customers and software developers.

Starting July 30, the artificial intelligence research laboratory reduced API rates for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%. Additionally, OpenAI debuted a "Fast mode" speed boost for its flagship GPT-5.6 Sol model, designed to process high-priority enterprise workloads up to 2.5 times faster than standard processing speeds.

The company said it is passing the reduced operational costs directly to developers, enterprise subscribers, and cloud platform partners like Amazon Web Services (AWS).

Retail sentiment on Stocktwits for OpenAI was ‘neutral’ with ‘normal’ message volumes. 

Infrastructure Efficiency Drives Steeper Price Cuts

The price adjustments alter the economics for high-volume automated AI workflows. Under the updated rate card, GPT-5.6 Luna costs $0.20 per million input tokens and $1.20 per million output tokens. Pricing for GPT-5.6 Terra has been set to $2 per million input tokens and $12 per million output tokens.

According to OpenAI, these price reductions extend across paid corporate accounts, including ChatGPT Work and Codex, where usage of Luna and Terra will consume fewer quota credits without changing monthly subscription tiers.

OpenAI attributes the cost drops to comprehensive optimizations throughout its software stack, including improved hardware routing, enhanced production inference software, and smarter context-caching algorithms that prevent AI agents from repeating completed computational work.

"Making advanced intelligence more abundant and affordable is central to OpenAI's mission," the company said in a statement, noting that Luna's reduced rate enables developers to run complex, multi-step agent loops economically at scale.

AI Models Are Turning More Efficient

In a notable development for autonomous software engineering, OpenAI revealed that the efficiency gains were partially generated by the AI models themselves.

During human-supervised testing, GPT-5.6 Sol autonomously rewrote and optimized critical production software kernels, dropping end-to-end model serving overhead by 20 percent. Furthermore, Sol conducted hundreds of automated generation experiments, improving token-generation efficiency by more than 15 percent.

These internal engineering breakthroughs allow OpenAI to extract higher computational output from existing server clusters, accelerating the feedback loop between model capability and serving efficiency.

For updates and corrections, email newsroom[at]stocktwits[dot]com.