Optimized for coding, computer automation, and subagent tasks. Delivers enterprise-scale efficiency with reduced latency for high-volume production workloads. GPT-5.4 mini is OpenAI's latest compact language model, delivering twice the speed of GPT-5 mini while maintaining superior performance. It supports multimodal inputs, tool use, function calling, and web search with a 400,000-token context window.

Optimized for high-volume production workloads, the model excels in multi-agent architectures where larger models handle planning while mini executes specialized subtasks efficiently.