An AI product team is debating when to use claude-3-opus versus claude-3-5-sonnet. A senior architect is asked to define the decision criteria. Which set of criteria correctly identifies use cases that genuinely benefit from Opus over Sonnet?

C is correct because Opus's primary differentiation is on tasks requiring deep, multi-step reasoning over ambiguous or complex inputs — research synthesis, complex architectural decisions, and novel problem solving where reasoning chain quality cascades to final answer quality. B is wrong because Sonnet and Opus share similar context window sizes; context length alone is not the distinguishing factor. A is wrong because Sonnet produces outputs that are indistinguishable from or superior to Opus on many common tasks — 'quality perception' is not a reliable criterion without task complexity analysis. D is wrong because classification tasks are typically well-suited to Haiku or Sonnet; they involve pattern matching over well-defined categories, not novel reasoning — using Opus wastes cost without quality gain.