Benchmark RadarAI BENCHMARK PROFILE
FM-Bench
FM-Bench evaluates LLM agents on long-horizon football club management through 20 in-game years with 26 tools, measuring managerial decision quality via a deterministic engine scoring.
- Released
- 2026-08-19
- Readiness
- Runnable
- Primary field
- Finance & Economics
Why it matters
Current agent benchmarks focus on bounded tasks, leaving long-horizon decision-making unmeasured. FM-Bench provides a reproducible environment with deterministic scoring to test sustained decision quality over hundreds of steps.
Motivation
Language model agents now execute bounded tasks reliably.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.