Omni-R1: benchmark results
Provider: Other.
Aggregate Rating 1582 ± 49, rank #1221 of 10916 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RAIL - Auditory Processing | 54.44 | LLM-as-judge score (%) on auditory-processing tasks (speech- | 96 |
| APUN-Bench (Synthetic) - Fine Pun Location | 49.43 | Fine pun-location accuracy (%), the predicted pun word exact | 88.9 |
| RAIL - Acquired Knowledge | 68.54 | LLM-as-judge score (%) on acquired-knowledge tasks over spee | 80 |
| RAIL - Memory | 73.19 | LLM-as-judge score (%) on auditory memory tasks (memory span | 80 |
| APUN-Bench (Synthetic) - Coarse Pun Location | 63.49 | Coarse pun-location accuracy (%), the predicted pun word mat | 77.8 |
| APUN-Bench (Synthetic) - Homographic Pun Inference | 39.73 | Homographic pun-inference accuracy (%), both predicted meani | 77.8 |
| RAIL - Fluid Reasoning | 58.39 | LLM-as-judge score (%) on fluid-reasoning tasks (induction, | 76 |
| APUN-Bench (Real-World) - Coarse Pun Location | 58.86 | Coarse pun-location accuracy (%), the predicted pun word mat | 66.7 |
| APUN-Bench (Real-World) - Fine Pun Location | 41.31 | Fine pun-location accuracy (%), the predicted pun word exact | 66.7 |
| APUN-Bench (Real-World) - Homographic Pun Inference | 35 | Homographic pun-inference accuracy (%), both predicted meani | 66.7 |
| APUN-Bench (Real-World) - Pun Recognition | 83.5 | Pun-recognition accuracy (%), deciding whether the audio cli | 66.7 |
| APUN-Bench (Synthetic) - Heterographic Pun Inference | 44.2 | Heterographic pun-inference accuracy (%), the predicted alte | 66.7 |
Interactive version: theaggregate.ai/model?slug=omni-r1 · How It Works · Data refreshed daily, snapshot 2026-11-08.