German Language LLM Index
a PeerBench project

Individual experiment result

GermEval NER (DE)

GLM-5.3 Flash Z.ai

Micro-F1 excluding MISC reasoning on / locked fp8
86.4%
Micro-F1 excluding MISC

Run details

Date
2026-08-30
Test cases
1,024
Median latency
0.00s
Total cost
Avg. prompt tokens
1,143.8
Avg. answer tokens
27
Avg. reasoning tokens
273.6
Quantization
fp8

About this benchmark

Native German named-entity recognition (person/location/org/misc) emitted as JSON; seqeval micro-F1 excluding MISC.

Named-entity recognition native · Native German Source ↗

Provenance

Canonical model
z-ai/glm-5.3-flash
Run ID
2026-08-30T16-35-15__openrouter__z-ai__glm-5.3-flash__germeval_de