Motif 3 (Beta): benchmark results
Provider: Motif. Released 2026-07-14. Access: Open.
Aggregate Rating 1670 ± 43, rank #457 of 2176 rated models, from 31 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Humanity's Last Exam | 40.36 | Accuracy (%) | 86.8 |
| Artificial Analysis Intelligence Index | 32.31 | Intelligence Index | 83 |
| AA GPQA Diamond | 86.87 | Accuracy (%) | 82.4 |
| AA-Omniscience Hallucination Rate | 49.1 | Hallucination Rate (%) | 79.4 |
| AA CritPt | 6 | Accuracy (%) | 72.5 |
| AA Long Context Reasoning | 74.67 | Accuracy (%) | 70.8 |
| AA Terminal-Bench 2.1 | 70.79 | Pass@1 (%) | 70.1 |
| Klu LLM Leaderboard | 66 | Klu Index | 69.6 |
| AA-Omniscience Index - Software Engineering (SWE) - PHP | -2 | Omniscience Index | 64.3 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 21 | Omniscience Index | 64.1 |
| AA-Omniscience Index - Software Engineering (SWE) - Python | -4 | Omniscience Index | 64.1 |
| AA-Omniscience Index - Science, Engineering & Mathematics | -7.51 | Omniscience Index | 63.9 |
Interactive version: theaggregate.ai/model?slug=motif-3-beta · How It Works · Data refreshed daily, snapshot 2026-10-28.