GPT-6 Luna vs Claude Fable 5.1

OpenAI's most efficient GPT-6 model, built for focused, high-volume work. Luna handles extraction, summarisation, classification, and other well-defined tasks that run thousands of times a day, at a price low enough to rival open models. It reads text and images, shares the family's very large context, and OpenAI reports it matches the previous flagship on factuality at a fraction of the cost. Its knowledge is the most recent in the family.

Anthropic's most capable generally available model, sitting above the Opus line. Fable 5.1 is built for long, unsupervised work: multi-day coding, agentic research, and tasks that run across dozens of tool calls. It reads text and images, maintains a very large context, and reasons before every answer at an adjustable depth. Its knowledge is the most recent of any Claude model. Anthropic positions Opus 5 as the starting point for most work, and Fable 5.1 for teams whose evaluations demand a higher ceiling.