Individual experiment result
MuSR (DE)
Gemini 3.1 Pro Google
Accuracy reasoning on / locked provider-internal
88.1%
Accuracy
Run details
- Date
- 2026-06-09
- Test cases
- 564
- Median latency
- 12.14s
- Total cost
- $10.49
- Avg. prompt tokens
- 1,464.9
- Avg. answer tokens
- 510.8
- Avg. reasoning tokens
- 795.7
- Quantization
- provider-internal
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 | 89.2% |
| object placements | 79.7% |
| team allocation | 89.2% |
Provenance
- Canonical model
- google/gemini-3.1-pro-preview
- Run ID
2026-06-09T18-30-10__vertex_ai__gemini-3.1-pro-preview__musr_de