OpenAI Launches GPT-5.4 Mini and Nano for Faster, Lower-Cost AI Workloads

OpenAI has expanded its artificial intelligence lineup with the release of GPT-5.4 mini and GPT-5.4 nano, two efficient models designed for fast, high-volume, latency-sensitive workloads.



The new models prioritize speed and cost efficiency over raw scale. GPT-5.4 mini runs more than twice as fast as GPT-5 mini while delivering improvements in coding, reasoning, multimodal understanding, and tool use. OpenAI says it approaches the performance of the full-sized GPT-5.4 model on several coding and computer use benchmarks. This makes it well suited for low-latency applications, including multimodal applications that reason over images in real time and analyzing dense user interface screenshots.


These capabilities tie into OpenAI's broader push around coding and multi-agent workflows. Following the rollout of the GPT-5.4 model featuring native computer control, the smaller mini variant is designed to work alongside it. In development environments like the Codex app for Mac, developers can set up multi-agent systems where a larger model handles planning while delegating repetitive tasks to GPT-5.4 mini subagents running in parallel. This approach helps reduce processing time and API costs for complex software projects.

For developers focused on minimizing cost, GPT-5.4 nano is the smallest and most affordable option in the lineup. OpenAI positions the nano model for classification, data extraction, ranking, and coding subagents that handle simpler supporting tasks.

GPT-5.4 mini is available starting today across the API, Codex, and ChatGPT. API pricing starts at $0.75 per million input tokens and $4.50 per million output tokens. It supports a 400,000-token context window, along with web search, file search, computer use, and skills. In Codex, the mini model uses just 30 percent of the standard GPT-5.4 quota. GPT-5.4 nano remains API-only, priced at $0.20 per million input tokens and $1.25 per million output tokens.

On the consumer side, GPT-5.4 mini is now available to free users and subscribers to the ChatGPT Go plan. It can be enabled through the "Thinking" feature in the + menu. For higher-tier plans, the mini model will automatically serve as a fallback when standard rate limits are reached.