German Language LLM Index
a PeerBench project

Individual experiment result

SB10K (DE)

DeepSeek V4 Flash 0731 DeepSeek

Matthews correlation coefficient reasoning off fp4
61.2%
Matthews correlation coefficient

Run details

Date
2026-08-11
Test cases
1,024
Median latency
0.67s
Total cost
$0.06
Avg. prompt tokens
778.6
Avg. answer tokens
2.7
Avg. reasoning tokens
Quantization
fp4

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
2026-08-11T11-46-52__openrouter__deepseek__deepseek-v4-flash-0731__sb10k_de