GPT 6 Luna

GPT-6 Luna is OpenAI's fast, cost-efficient GPT-6 model for high-volume chat, classification and lightweight agentic tasks. It accepts text and image inputs and retains the series' 1.05-million-token context and 128,000-token output limit.

Luna brings the GPT-6 generation to workloads where inference expenditure constrains repetition. OpenAI reports improvements in professional automation, factual reliability and software engineering over its predecessor.

A useful way to evaluate an economical model is to measure accepted results, including retries and human corrections. Classification, extraction and repeated drafting each need different acceptance criteria; the cheapest individual response need not produce the cheapest completed workflow.

The announcement also describes clearer communication and fewer misleading claims about coding work. Those improvements support collaboration, but they do not replace verification: keep concrete checks around any output that another process will consume.