Individual experiment result
GermEval NER (DE)
GPT-5.5 OpenAI
Micro-F1 excluding MISC reasoning on / locked provider-internal
85.9%
Micro-F1 excluding MISC
Run details
- Date
- 2026-06-13
- Test cases
- 1,024
- Median latency
- 1.86s
- Total cost
- $7.08
- Avg. prompt tokens
- 1,140.7
- Avg. answer tokens
- 29.1
- Avg. reasoning tokens
- 11.3
- Quantization
- provider-internal
About this benchmark
Native German named-entity recognition (person/location/org/misc) emitted as JSON; seqeval micro-F1 excluding MISC.
Provenance
- Canonical model
- openai/gpt-5.5
- Run ID
2026-06-13T11-17-57__openai__codex__gpt-5.5-low__germeval_de