German Language LLM Index
a PeerBench project

Individual experiment result

SB10K (DE)

Gemini 3.8 Flash Google

Matthews correlation coefficient reasoning on / locked provider-internal
66.9%
Matthews correlation coefficient

Run details

Date
2026-09-04
Test cases
1,024
Median latency
1.64s
Total cost
$1.88
Avg. prompt tokens
625.7
Avg. answer tokens
1.7
Avg. reasoning tokens
118.3
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

Run ID
2026-09-04T15-03-30__vertex_ai__gemini-3.8-flash__sb10k_de