Verdict
Built from the measurements below
- Claude Opus 5.5 is more accurate: 85.0% against 68.8% on the same 80 questions, a 16.2-point gap that is statistically significant.
- Claude Haiku 5.5 is 31× cheaper per question as billed ($0.00048 vs $0.015), reasoning tokens included.
- Claude Haiku 5.5 generates text 1.5× faster: 192 tokens per second against 129.
- Claude Haiku 5.5 starts answering sooner: 0.7 s to the first token, against 2.3 s for Claude Opus 5.5.
Which should you choose?
Choose Claude Haiku 5.5 for the lower cost per correct answer ($0.0007 against $0.017), faster streaming and a quicker first token. Choose Claude Opus 5.5 for the most accurate answers of the two. The widest subject gap is in law: Claude Opus 5.5 answered 5 of 6 correctly, Claude Haiku 5.5 2 of 6. With so few questions per subject, treat this as a lead to test on your own prompts, not a result.
Accuracy gap -16.2 points (z = -2.44). We call a winner only when z is at least 1.96, the usual 95% threshold. Speed, latency and price need a 10% margin.