German Language LLM Index
a PeerBench project

Individual experiment result

ScaLA (DE)

Qwen3.5 9B Alibaba

Matthews correlation coefficient reasoning off bf16
62.4%
Matthews correlation coefficient

Run details

Date
2026-06-10
Test cases
2,048
Median latency
0.82s
Total cost
$0.03
Avg. prompt tokens
928.9
Avg. answer tokens
2.5
Avg. reasoning tokens
Quantization
bf16

About this benchmark

Native German linguistic acceptability — is the sentence grammatical (ja/nein)? generate_until; primary metric MCC (accuracy recorded as secondary).

Binary acceptability native · Native German Source ↗

Provenance

Canonical model
qwen/qwen3.5-9b
Run ID
2026-06-10T12-29-47__openrouter__qwen__qwen3.5-9b__scala_de