P

rubberband · Production library run 1

67.5Library score
Pass rate92.5%
Simplicity79.6
Problems $$11.89

Implementers use the real production library (tantivy).

ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE94.3%77.6$1.88
69.0
GLM 5.3 Flash · mini-SWE94.0%79.0$7.08
69.0
GPT-5.6 Luna · Codex88.9%82.4$2.93
64.4
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
case-law-export51.649.147.349.4session-recall49.637.692.059.7document-finder57.053.566.759.1code-chunk-search55.050.542.449.3symbol-resolution42.560.450.951.3docs-search59.861.260.960.7knowledge-ingest89.192.577.186.2federated-search64.061.797.474.4post-indexer87.686.685.286.5blog-rebuild73.133.060.155.4mailbox-tags84.991.094.290.1attachment-explorer74.669.175.373.0archive-search76.191.680.982.9index-admin96.691.60.262.8file-locator87.882.195.788.5project-rule-search76.667.871.071.8blog-filters46.968.642.652.7dictionary-lookup63.076.762.167.2archive-timeline75.782.085.481.0source-rule-search68.774.10.047.6