Model
160d agoGPT‑5.4 Mini and Nano Launch
OpenAI unveils GPT‑5.4 mini and nano, lightweight variants that run faster and use less memory. Designed for coding, tool integration, and multimodal reasoning, they excel in high‑volume API calls and sub‑agent tasks. The models maintain GPT‑5.4’s performance while cutting inference time, making them ideal for production workloads.
The Signal
They also support rapid prototyping and low‑latency inference across edge devices.