ERPBench (Market Ecologies) - Solo: leaderboard
ERPBench (2026) evaluates enterprise decision agents in a six-round ERP simulation of a cereal manufacturer with coupled pricing, production, procurement, inventory and finance decisions over 100 fixed problems, repairing or clamping infeasible actions. This board is the Solo ecology, where each agent competes against fixed rule-based opponents; the metric is mean sixth-round company valuation in millions, an unbounded monetary outcome, higher is better. It is a different benchmark from ERP-Bench and from the ERPNext-based ERPBench.
Metric: Mean terminal company valuation (M; sixth-round valuation over 100 ERP problems, each agent against fixed rule-based opponents). Source: arxiv.org. Saturation forecast: Not forecast. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Flash | 252.29 | #254 |
| 2 | Qwen 3.7 Max | 205.87 | #142 |
| 3 | GPT-5.5 | 186.43 | #48 |
| 4 | Seed 2.0 Pro | 181.27 | #209 |
| 5 | Claude Opus 4.6 | 176.78 | #100 |
| 6 | Gemini 3.1 Pro (Preview) | 169.41 | #91 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=erpbench-market-ecologies-solo · How It Works · Data refreshed daily, snapshot 2026-11-12.