Overview
On July 30, 2026, OpenAI announced sweeping price reductions across two tiers of its GPT-5.6 large language model (LLM) family, cutting GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, just three weeks after the models reached general availability. The cuts were disclosed in a post on OpenAI's official website and confirmed across the company's application programming interface (API) pricing documentation.
Pricing Details
Following the July 30 adjustment, <cite index="1-1">API pricing is $2 per million input tokens and $12 per million output tokens for Terra, and $0.20 per million input tokens and $1.20 per million output tokens for Luna.</cite> Luna's pre-cut rates had stood at <cite index="7-16">$1.00 input and $6.00 output at launch; Terra was $2.50/$15.00.</cite> <cite index="3-6">The flagship Sol model remained unchanged at $5/$30 per million tokens.</cite>
For context on generational trajectory, <cite index="2-10">roughly four months after its release, OpenAI is selling March's full flagship intelligence at about one-thirteenth the token price.</cite>
Model Architecture
<cite index="20-5">OpenAI launched GPT-5.6 as a family comprising Sol, its new flagship; Terra, a balanced model for everyday work; and Luna, its most cost-efficient model.</cite> <cite index="22-6">OpenAI explains that "the number identifies a model's generation, while Sol, Terra, and Luna identify durable capability tiers that can advance on their own cadence."</cite> <cite index="5-7,5-8">Sol is aimed at the most complex reasoning-heavy and agentic workloads, including advanced coding and multi-step planning, while Luna is positioned for high-throughput, low-latency tasks such as summarization, classification, routing, and lightweight real-time assistants where cost per request is the primary constraint.</cite>
<cite index="8-8">The GPT-5.6 family reached general availability on July 9, 2026, with a 1.05 million-token context window on all three tiers.</cite> <cite index="27-1,27-2">OpenAI initially limited GPT-5.6 to approximately 20 U.S.-government-approved organizations at Washington's request, citing the model's advanced cybersecurity capabilities; the Commerce Department cleared broad release two weeks later, on July 9, 2026.</cite>
Stated Rationale
<cite index="14-7">OpenAI attributed the reductions to efficiency improvements made during internal development of GPT-5.6, including the model's role in optimizing production software and improving speculative decoding.</cite> <cite index="9-12">OpenAI says its inference work cut end-to-end serving costs by 20% and improved token-generation efficiency by more than 15%.</cite> Chief Executive Officer Sam Altman framed the move in competitive terms, writing on X that <cite index="15-4">"We want to offer the best price/intelligence tradeoff at every level."</cite>
Competitive Context
<cite index="11-2">OpenAI's price cuts may intensify competition in the industry as U.S. companies battle cheaper Chinese rivals for customers increasingly wary of the technology's ballooning costs.</cite> <cite index="3-9">A CNBC investigation published on July 7, 2026, revealed that Chinese models have captured 46% of U.S. enterprise token usage on OpenRouter, at times peaking above U.S.-origin models.</cite> <cite index="3-11">DeepSeek V4 Pro, for instance, is priced at $0.435/$0.87 per million tokens, benefiting from a standing 75% promotional discount.</cite>
<cite index="5-3">The cuts arrive just a few days after Anthropic released its highly performant Claude Opus 5 at the same price as Opus 4.8, and Google introduced Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, two rival models built around lower inference costs, faster execution and more efficient agent workloads.</cite> <cite index="5-6">OpenAI is directly cutting per-token rates, Google is pairing lower prices with reductions in token use and tool calls, and Anthropic is emphasizing stronger task performance for the same price.</cite>
Financial Backdrop
<cite index="14-2">OpenAI disclosed that its models now reach more than one billion active users and more than two million businesses, an announcement that followed the price reductions.</cite> However, the company continues to operate at a loss: <cite index="14-8,14-9">OpenAI is operating at a loss as it continues to spend on model development and infrastructure, with financials showing revenue of $13.07 billion in 2025 against a net loss of $38.5 billion.</cite> <cite index="16-13">OpenAI has committed to some $600 billion in compute spend by 2030.</cite>
<cite index="1-2">GPT-5.6 Terra and Luna remain available in ChatGPT Work, Codex, and the OpenAI API.</cite> <cite index="1-7">Pricing changes also began rolling out on Amazon Web Services (AWS) on July 30.</cite>