Loading model…
anthropic/claude-opus-4.813/25 in v2 — Resilience, score 69.7%, pass rate 60.1% (276/459)1.66s, output speed 187.1 tok/s$9.39, prompt tokens 45972, completion tokens 37027996.7% pass rate (58/60)96.7% pass rate (29/30)88.9% pass rate (8/9)21.7% pass rate (13/60)13.3% pass rate (8/60)40.0% pass rate (24/60)openai/gpt-5.5 (score 77.9%, pass rate 61.9%), openai/gpt-5.3-codex (score 77.7%, pass rate 61.2%), and qwen/qwen3.7-max (score 75.3%, pass rate 54.9%)Claude Opus 4.8 ranks 13/25 in Resilience, showing solid coding and cost efficiency but significant gaps in debugging, refactoring, and reasoning. Suitable for deployment where cost matters and reasoning demands are moderate, but not ideal for complex debugging or refactoring tasks.
Overall score % from merged run_models rows (chronological). Only runs that include this model appear as points.
Score % vs pass rate % per category. With 0/1 scorers, both usually line up; with proportional tests, score % reflects partial credit while pass rate counts tests that clear the fixture threshold.
Total estimated spend per scope for this model (bars, left axis) and mean spend per merged result row (line, right axis: total ÷ tests).
Pass rate % per difficulty level — complements the score % view above.
Normalized 0–100 within this model: TTFT (shorter → higher spoke) and decode tok/s (higher → higher spoke). Values come from streamed BLXBench runs merged into overall_ranking.json.
Pass rate % per category for this model (distinct from score %, which reflects partial credit).