Hugging Face Blog·· 2026-07-01
ScarfBench: Benchmarking AI Agents for Enterprise Java Framework Migration
ScarfBench: Benchmarking AI Agents for Enterprise Java Framework Migration
AI summary
Hugging Face released ScarfBench (Self-Contained Application Refactoring Benchmark) to evaluate AI agents migrating enterprise Java applications between Spring, Jakarta EE and Quarkus.
Selection record
Threshold 60Official, first-handFirst 47Second 47
Not admittedSum of both 94 < twice the threshold 120
- Source tier
- Official, first-hand; this tier's threshold is 60
- Pre-filter
- passed：AI Agent企业Java迁移基准评测
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Hugging Face Blog · huggingface.co