.jpg)
OpenAI has introduced its latest lightweight AI models—GPT-5.4 mini and GPT-5.4 nano positioning them as faster and more cost-efficient options for developers, particularly in coding and high-volume use cases.
In a blog post announcing the release, the company described the models as its “most capable small models yet,” bringing several strengths of its larger GPT-5.4 system into more efficient formats designed for speed and scalability.
According to OpenAI, GPT-5.4 mini significantly outperforms its predecessor, GPT-5 mini, across areas such as coding, reasoning, multimodal understanding and tool usage, while operating at more than twice the speed. The company added that in certain evaluations, the model approaches the performance levels of the full GPT-5.4 model.
The GPT-5.4 nano variant, meanwhile, has been positioned as the smallest and most affordable option in the lineup. It is designed for tasks that require high efficiency, including classification, data extraction, ranking and lower-level coding support.
OpenAI emphasised that both models are optimised for coding workflows, enabling use cases such as targeted code edits, navigating large codebases, generating front-end components and debugging loops with minimal latency.
The models are tailored for applications where responsiveness is critical, including coding assistants, AI subagents, computer-use systems that interpret screenshots, and real-time multimodal tools. The company noted that in such environments, faster and more efficient models can often outperform larger systems in delivering a seamless user experience.
GPT-5.4 mini is available across API, Codex and ChatGPT, while GPT-5.4 nano is currently limited to API access. The nano model is priced at $0.20 per one million input tokens and $1.25 per one million output tokens, reflecting its positioning as a cost-effective solution for large-scale deployments.
