Claude Sonnet 5.5 matches Opus 5.5 on coding tests at half the price
AIOpenRouter tested Claude Opus 5.5, Sonnet 5.5, and Haiku 5.5 on 11 coding tasks with hidden tests and all three passed every one. Sonnet 5.5 scored higher than Opus 5.5 on GPQA Diamond (92.4% vs 90.4%) and τ-bench airline (78.7% vs 75.5%) at half the per-token price, while Haiku 5.5 cost about a 19th to a 30th as much as Opus per coding task.
Source details
OpenRouter Blog · openrouter.aiPublished · added here
Original title: Claude Opus vs Sonnet vs Haiku: Which Should You Use?