Benchmark RadarAI BENCHMARK PROFILE
DocFormBench
DocFormBench evaluates content-aware document formatting for LLMs and multimodal models, using accuracy and efficiency metrics, with a workflow method DocFormFlow for target localization and modification.
- Released
- 2026-06-01
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Existing formatting benchmarks lack content-aware evaluation, leaving a gap in assessing target identification. This benchmark addresses that by providing diverse formatting requirements and metrics for accuracy and efficiency, enabling practical comparison of formatting models.
Motivation
Recent advances in large language models (LLMs) have opened up new possibilities for automated document formatting.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.