OpenAI’s Latest Update Brings Efficiency Gains to its AI Models, Reducing Costs for Customers
In a significant move, OpenAI has announced that it has reduced the prices of two of its advanced language models, GPT-5.6 Luna and Terra, by as much as 80% and 20%, respectively. This development is aimed at making its models more cost-efficient, allowing customers to access cutting-edge AI capabilities without breaking the bank.
The updated pricing structure means that GPT-5.6 Luna now costs $0.20 per million input tokens and $1.20 per million output tokens, down from $1 and $6 previously. Similarly, Terra has seen its prices drop from $2.50 to $2 per million input tokens and from $15 to $12 per million output tokens. This reduction in cost is expected to benefit customers who use these models for various tasks, such as content generation, language translation, and coding assistance.
OpenAI’s decision to reduce costs also affects how usage is counted in its Codex and ChatGPT Work platforms. For instance, if new tasks use the GPT-5.6 Luna model, they will deduct less from customers’ allowances, allowing them to complete more work under the same quota. Furthermore, OpenAI has upgraded Auto-review in the ChatGPT app and Codex CLI from GPT-5.4 to GPT-5.6 Luna, which should reduce costs by approximately ten times.
In addition to these efficiency gains, OpenAI has also introduced a “Fast mode” for API customers using its GPT-5.6 Sol model. This feature promises speeds up to 2.5 times faster than standard processing without compromising the model’s intelligence. However, it comes at twice the standard API price, making it suitable primarily for time-sensitive applications such as coding, research, and agentic workloads.
The reduction in costs is attributed to OpenAI’s recent improvements to its GPT-5.6 Sol model, which have enabled efficiency gains that allow for lower prices without sacrificing performance. In fact, the company claims that GPT-5.6 Luna tops its intelligence index among compared models, despite being significantly cheaper per task.
As customers increasingly rely on AI-powered tools and services, it is essential to remember that even with reduced costs, these models are not immune to security risks. Organizations must continue to prioritize security measures and regularly test their systems to prevent potential breaches.
In light of OpenAI’s latest update, customers can take advantage of the efficiency gains by opting for GPT-5.6 Luna or Terra, which offer significant cost savings without compromising performance. Additionally, businesses should remain vigilant about security risks associated with AI-powered tools and services, and regularly review their systems to prevent potential breaches.
Source: Bleeping Computer — 2026-07-31