83.7%
Accuracy
Run details
- Date
- 2026-06-10
- Test cases
- 564
- Median latency
- —
- Total cost
- —
- Avg. prompt tokens
- —
- Avg. answer tokens
- —
- Avg. reasoning tokens
- —
- Quantization
- bf16
About this benchmark
German MuSR v1.1 defect-filtered benchmark: 553 questions (11 truncated team-allocation translations excluded), generate_until cot+; primary metric accuracy.
Subject breakdown
| Subject | Score |
|---|---|
| murder mystery | 85.2% |
| object placements | 73.4% |
| team allocation | 84.8% |
Provenance
- Canonical model
- google/gemma-4-26b-a4b-it
- Run ID
musr_de