LongCat 2.0 — benchmark results
Meituan's open-weights 1.6T-A48B mixture-of-experts model with native 1M context via LongCat Sparse Attention. Provider: Meituan. Released 2026-07-05. Access: Open.
Unified ELO 1704 ± 22, rank #273 of 1841 rated models, from 37 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Humanity's Last Exam | 32.07 | Accuracy (%) | 90.4 |
| AA Omniscience - Science, Engineering & Mathematics | 38.7 | Accuracy (%) | 81.8 |
| AA Omniscience - Software Engineering (SWE) - Go | 38 | Accuracy (%) | 81.2 |
| Artificial Analysis Intelligence Index | 33.48 | Intelligence Index | 79.4 |
| AA Omniscience - Health | 29.7 | Accuracy (%) | 79.3 |
| AA Omniscience - Business | 25.7 | Accuracy (%) | 78.6 |
| AA CritPt | 2.57 | Accuracy (%) | 78.4 |
| AA-Omniscience Accuracy | 29.8 | Accuracy (%) | 77.6 |
| AA Omniscience - Software Engineering (SWE) - Kotlin | 34 | Accuracy (%) | 77.5 |
| AA Omniscience - Software Engineering (SWE) - JavaScript | 46.36 | Accuracy (%) | 76.2 |
| AA Omniscience - Software Engineering (SWE) - TypeScript | 42.22 | Accuracy (%) | 75.1 |
| AA Omniscience - Software Engineering (SWE) - Julia | 28 | Accuracy (%) | 74.8 |
Interactive version: theaggregate.ai/model?slug=longcat-2-0 · How It Works · Data refreshed daily, snapshot 2026-07-25.