In a significant move in the artificial intelligence space, OpenAI has introduced two new models—GPT-5.4 Mini and GPT-5.4 Nano. These models bring many of the capabilities of the flagship GPT-5.4 into more efficient, cost-effective versions designed for speed and scalability.
More Power at Lower Cost
GPT-5.4 Mini is a major upgrade over its predecessor, offering improved performance in coding, reasoning, and multimodal understanding, including both text and image inputs. Despite being smaller, it delivers near GPT-5.4-level results while operating at significantly lower cost and nearly double the speed, making it ideal for developers and businesses looking to optimize performance without increasing expenses.
Nano Model Focused on Speed and Efficiency
The GPT-5.4 Nano model is specifically designed for tasks where speed and affordability are critical. These include data classification, data extraction, ranking, and lightweight coding tasks. Its streamlined architecture makes it highly efficient for handling large volumes of smaller, repetitive operations.
Built for High-Speed AI Systems
Both models are well-suited for systems that require rapid processing, such as coding assistants, AI agents handling multiple micro-tasks, screenshot analysis tools, and real-time image processing applications.
Smarter Workflows with Codex
On platforms like Codex, the flagship GPT-5.4 model can handle high-level planning, while GPT-5.4 Mini takes care of smaller tasks. This division of work improves overall speed and can reduce operational costs by up to one-third, as the Mini model uses only about 30% of the resources.
Available Across Platforms
GPT-5.4 Mini is now available on ChatGPT, via API, and in Codex. It supports both text and image inputs and includes features like web browsing, file search, and computer-based task execution. Users on Free and Go plans can access it through the “Thinking” option.
Nano Model for Developers
GPT-5.4 Nano is currently available only through API access, allowing developers to integrate it into their apps and services.
Pricing Details
In terms of pricing, GPT-5.4 Mini costs approximately $0.75 per million input tokens and $4.50 per million output tokens. The Nano model is even more affordable, priced at around $0.20 per million input tokens and $1.25 per million output tokens.
With these launches, OpenAI aims to make advanced AI more accessible, efficient, and scalable—empowering developers and businesses to build faster and smarter applications.
