FrontierCode’s design explains why Opus 5 scores drop as reasoning increases
silasalberti · x · 2026-07-25
The author explains why Opus 5’s FrontierCode scores can fall as reasoning effort rises.
FrontierCode is designed to measure merge-ability, not just correctness. It includes a scope criterion that penalizes edits beyond what the task actually requires. The same pattern appears across all frontier models tested, but it is most pronounced in Opus 5, which becomes more likely to refactor code unprompted at higher reasoning levels.
Related event: Opus 5 Coding Paradox: Higher Reasoning Leads to Lower Scores(13 posts)→
More from Models
- BullshitBench update: GPT-6-Astra beats all prior OpenAI models but still trails Anthropic — scaling01 · 2026-09-11
- Astra Scores 83% on GauntletBench, First Computer-Use Agent to Beat Human Baseline — ducha_aiki · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- DeepSeek V4 Pro API to continue after Sept 2026, billing unchanged — teortaxesTex · 2026-09-11
- DeepSeek V4.1 Flash Hits 98% of GPT-6 Astra's Score at 1.4% of the Cost in Third-Party Benchmark — ayushtweetshere · 2026-09-11