Skip to content

Benchmark

E-Commerce Bench · BadSpend · fraudulent-supplier spend

Scores are percentages. Lower is better.

Models and systems measured
20
Scores recorded
21
Sources
1
IQ weight
0.10

Score distribution

11.6%
Mean
10.7%
Median
7.3%
Std dev
2.1%
10th percentile
19.7%
90th percentile
0.12%
Min
0.12%27.9%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.10 for chat models, measured over 17 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or systemScore
1Claude Opus 4.7benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs0.12%
2GLM 5.2benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · high reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs2%
3Gemini 3.5 Flashbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs2.1%
4Claude Fable 5benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs3.5%
5Claude Opus 4.8benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs5.4%
6Kimi K3benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs5.9%
7GLM 5.1benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs6.5%
8Qwen3.7 Maxbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.9.8%
9Qwen3.6 Plusbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.10.7%
10Gemini 3.1 Pro Previewbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs12.6%
11Claude Opus 4.6benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs15%
12GPT 5.5benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs16.6%
13GLM 5.3benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs16.7%
14Kimi K2.6benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs17.2%
15Qwen3.8 Maxbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.18.4%
16GPT-5.6 Solbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs18.5%
17GLM 5.2benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs19.7%
18GLM 5.3 Flashbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs27.9%

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
DeepSeek V4 Pro Preview (E-Commerce Bench)benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · max reasoning · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs9%
Qwen3.5 Plusbenchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.20.1%
Qwen3.8 Max Preview (E-Commerce Bench)benchmark-owner 365-day merchant simulation · five independent episodes · up to four stores · initial balance ¥100,000 · thinking enabled where supported; effort not specified · exact API snapshot not stated for unversioned labels· Reported by E-Commerce Bench authors / Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.6.1%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.