Anthropic says Claude Opus 5.5 performs at the level of Claude Fable 5.1, its most intelligent model, for most tasks, while being faster and more efficient than Opus 5. The company also says Opus 5.5 writes more clearly and puts the most important information first, which it says makes the model easier to work with over long sessions. The reason that matters for long tasks is practical: a model working for a long time produces a lot of output, and leading with the key point lets a person decide quickly whether to act or intervene. All of this is Anthropic's own assessment, and no independent comparison is cited. Launch video description