Ultiima-72B — benchmark results
Sakalti's community TIES merge of Qwen 2.5 72B fine-tunes. Provider: Other. Released 2025-01-09. Access: Open.
Unified ELO 1636 ± 10, rank #398 of 1776 rated models, from 129 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - GPQA | 21.92 | Score | 99.7 |
| Open LLM Leaderboard - BBH | 61.1 | Score | 99.6 |
| Open LLM Leaderboard - MMLU-Pro | 54.51 | Score | 99.6 |
| Open Arabic LLM - Arabic MMLU Computer Science (High School) | 82.76 | Accuracy (%) | 99.4 |
| Open Arabic LLM - Arabic MMLU HT Prehistory | 81.17 | Accuracy (%) | 99.4 |
| Open Arabic LLM - Alghafa Multiple Choice Grounded Statement Xglue Mlqa Task | 94.67 | Accuracy (%) | 99.1 |
| Open Arabic LLM - Arabic MMLU General Knowledge | 73.61 | Accuracy (%) | 98.8 |
| Open Arabic LLM - Arabic MMLU HT College Medicine | 73.99 | Accuracy (%) | 98.8 |
| Open Arabic LLM - Aratrust PhysicalHealth | 94.52 | Accuracy (%) | 98.8 |
| Open Arabic LLM - Arabic MMLU General Knowledge (Middle School) | 83.72 | Accuracy (%) | 98.5 |
| Open Arabic LLM - Arabic MMLU HT Human Aging | 75.34 | Accuracy (%) | 98.5 |
| Open Arabic LLM - Arabic MMLU Physics (High School) | 69.02 | Accuracy (%) | 98.5 |
Interactive version: theaggregate.ai/model?slug=ultiima-72b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.