German Language LLM Index
a PeerBench project

Individual experiment result

SB10K (DE)

Gemini 4 Argon Google

Matthews correlation coefficient reasoning off provider-internal
71.8%
Matthews correlation coefficient

Run details

Date
2026-10-13
Test cases
1,024
Median latency
1.54s
Total cost
$1.58
Avg. prompt tokens
748.9
Avg. answer tokens
4.3
Avg. reasoning tokens
—
Quantization
provider-internal

About this benchmark

Native German Twitter sentiment (positive/neutral/negative), generate_until variant; primary metric MCC.

3-class sentiment native · Native German Source ↗

Provenance

Canonical model
google/gemini-4-argon
Run ID
2026-10-13T19-10-37__vertex_ai__gemini-4-argon__sb10k_de