Ling-1T — benchmark results
Ant Group inclusionAI's open trillion-parameter MoE flagship (~50B active), tuned for code, math and efficient inference (October 2025). Provider: InclusionAI. Released 2025-10-09. Access: Open.
Unified ELO 1576 ± 15, rank #544 of 1776 rated models, from 61 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA MMLU-Pro | 82.2 | Accuracy (%) | 79.1 |
| UGI - Natural Intelligence | 30.84 | NatInt Score | 76.5 |
| AA LiveCodeBench | 67.72 | Pass@1 (%) | 75.4 |
| CritPt | 0.6 | Accuracy (self-reported) | 71.1 |
| UGI Leaderboard | 41.3 | UGI Score | 70.5 |
| AA Omniscience - Software Engineering (SWE) - HTML | 40 | Accuracy (%) | 67.5 |
| AA AIME 2025 | 71.33 | Accuracy (%) | 65.6 |
| AA CritPt | 0.57 | Accuracy (%) | 64.9 |
| AA Omniscience - Software Engineering (SWE) - Java | 21 | Accuracy (%) | 63.8 |
| AA Omniscience - Software Engineering (SWE) - Python | 24.5 | Accuracy (%) | 60.8 |
| OpenCompass Language - Dialogue | 96.3 | Score (%) | 60 |
| AA Omniscience - Software Engineering (SWE) - Kotlin | 22 | Accuracy (%) | 59.9 |
Interactive version: theaggregate.ai/model?slug=ling-1t · How the rankings work · Data refreshed daily, snapshot 2026-07-22.