OpenAI’s new GPT-5.4 Mini and Nano models are officially here, redefining efficiency.

Mini packs massive coding power at 54.4% on SWE-Bench, while Nano slashes costs to just $0.20 per million tokens.

Designed for speed, these subagents handle ranking and data tasks, allowing developers to build lightning-fast, multi-model AI architectures.

The future of scalable, cost-effective artificial intelligence has finally arrived.