ConfBench - Error Capture (ECARB@30): leaderboard

Metric: Error capture at a 30 percent review budget (ECARB; errors found by reviewing the least-confident 30 percent of predictions divided by the errors a random 30 percent review finds, so 1 is no lift and 3.33 the ceiling; verbalized top-4 guesses with probabilities in one call, OCR text with Amazon Textract confidences plus the document image, over the entity fields of 1,346 degraded variants of 75 RealKIE-FCC-Verified invoices). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.

Top models

#ModelScore
1Claude Opus 4.62.43
2Claude Sonnet 4.52.24
3Claude Haiku 4.51.96
4Qwen 3.6 27B1.93
5Kimi K2.51.68
6Qwen 3 VL 235B A22B1.53
7Gemma 3 12B1.39

Interactive version: theaggregate.ai/benchmark?slug=confbench-error-capture-ecarb-30 · How It Works · Data refreshed daily, snapshot 2026-09-29.