OpenAI has sharply cut prices across its GPT-5.6 frontier model series, reducing the cost of GPT-5.6 Luna, the smallest and fastest model in the lineup, by 80 percent, while cutting the mid-tier Terra model by 20 percent. Pricing for the flagship Sol model remained unchanged, though OpenAI introduced a premium Fast mode for Sol that runs roughly 2.5 times quicker at double the price.
Under the new pricing, Luna now costs $0.20 per million input tokens and $1.20 per million output tokens, down from $1 and $6 respectively, bringing its combined input-output price from $7 to $1.40 per million tokens. Terra falls to $2 per million input tokens and $12 per million output tokens, down from $2.50 and $15. The changes took effect immediately across the API, ChatGPT, and Codex.
The reduction places Luna below Google’s Gemini 3.5 Flash-Lite, which costs a combined $2.80 per million tokens, and far below Gemini 3.6 Flash at $9, while also undercutting DeepSeek on input costs, though it remains more expensive on output tokens. OpenAI attributed the cuts to efficiency gains, saying its Sol model had been used to rewrite and optimize its own inference infrastructure, autonomously improving GPU utilization and production code, resulting in a 20 percent reduction in end-to-end serving costs and a more than 15 percent improvement in token generation throughput through speculative decoding. The company said tasks that previously cost $1 on Luna now run for approximately six cents at nearly nine times the original speed, positioning Luna for high-volume work such as large-scale document analysis, customer interaction sorting, and routine coding tasks within agent workflows.
The cuts arrive just weeks after OpenAI launched the GPT-5.6 series on July 9, and days after Anthropic released Claude Opus 5 at the same price as its predecessor and Google introduced Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, both built around lower inference costs. The move comes as ChatGPT has surpassed one billion monthly active users, and as enterprise customers increasingly push back on AI costs, with several companies citing meaningful savings from the new pricing in their own deployments. Anthropic’s introductory pricing for Claude Sonnet 5 is set to expire August 31, a date that could prompt further competitive repricing across the industry as AI providers continue shifting their competitive focus from raw model capability toward cost and operational efficiency.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
