German Language LLM Index
a PeerBench project

Individual experiment result

ScaLA (DE)

Gemma 4 12B Google

Matthews correlation coefficient reasoning off bf16
64.1%
Matthews correlation coefficient

Run details

Date
2026-06-11
Test cases
2,048
Median latency
27.85s
Total cost
Avg. prompt tokens
910.7
Avg. answer tokens
2.6
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
google/gemma-4-12b-it
Run ID
2026-06-11T14-44-14__openai__gemma4-12b-bf16__scala_de