80.9%
Accuracy
Run details
- Date
- 2026-06-11
- Test cases
- 564
- Median latency
- 61.14s
- Total cost
- $0.22
- Avg. prompt tokens
- 1,469.2
- Avg. answer tokens
- 1,629.7
- 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 | 77.6% |
| object placements | 73.4% |
| team allocation | 86.0% |
Provenance
- Canonical model
- qwen/qwen3.5-9b
- Run ID
2026-06-11T14-33-49__openai__qwen35-9b-bf16__musr_de