Benchmark
Roomote Role Evals
Scores are percentages.
Benchmark sourceScores from Roomote Role Evals data.
- Models measured
- 21
- Scores recorded
- 115
- Sources
- 1
- IQ weight
- 1.68
Score distribution
- 65.7%
- Mean
- 73%
- Median
- 28.2%
- Std dev
- 27%
- 10th percentile
- 96%
- 90th percentile
- 0%
- Min
This benchmark has a weight of 1.68 for chat models, measured over 21 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 101 | Grok 4.6review· Roomote Role Evals data | 33% |
| 102 | Gemini 3.6 Flashreview· Roomote Role Evals data | 32% |
| 103 | Inkling FP4vision-extraction· Roomote Role Evals data | 29% |
| 104 | DeepSeek V4 Pro 0813review· Roomote Role Evals data | 27% |
| 105 | DeepSeek V4 Proreview· Roomote Role Evals data | 24% |
| 106 | Inkling FP4advisor· Roomote Role Evals data | 23% |
| 107 | Kimi K2.7 Codereview· Roomote Role Evals data | 8% |
| 108 | MiniMax M3review· Roomote Role Evals data | 7% |
| 109 | DeepSeek V4 Flash 0731vision-extraction· Roomote Role Evals data | 0% |
| 110 | DeepSeek V4 Provision-extraction· Roomote Role Evals data | 0% |
| 111 | DeepSeek V4 Pro 0813vision-extraction· Roomote Role Evals data | 0% |
| 112 | Doubao Seed 2.0 Codereview· Roomote Role Evals data | 0% |
| 113 | Doubao Seed 2.0 Codevision-extraction· Roomote Role Evals data | 0% |
| 114 | GLM 5.2vision-extraction· Roomote Role Evals data | 0% |
| 115 | Inkling FP4review· Roomote Role Evals data | 0% |
Showing 101–115 of 115 results.