Kimi K2.6 (Non-reasoning): benchmark results
Kimi K2.6 evaluated with reasoning disabled. Provider: Moonshot. Released 2026-04-20. Access: Open.
Aggregate Rating 1661 ± 10, rank #487 of 2168 rated models, from 150 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RuleWeaver - Same-Source - Rule Recall | 72.92 | Gold-rule recall (%) | 100 |
| SABRE-Prior - Texture | 52 | Strict All4 accuracy (%; a case counts only when all four ye | 100 |
| AA Omniscience - Software Engineering (SWE) - Rust | 74 | Accuracy (%) | 95.1 |
| UGI - Writing | 64.19 | Writing Score | 94 |
| AA Omniscience - Software Engineering (SWE) - Go | 40 | Accuracy (%) | 93.2 |
| AA Omniscience - Software Engineering (SWE) - PHP | 48 | Accuracy (%) | 92.3 |
| UGI - Natural Intelligence | 59.46 | NatInt Score | 92 |
| AA Omniscience - Software Engineering (SWE) - Kotlin | 40 | Accuracy (%) | 91.6 |
| AA TAU-2 Bench | 93.86 | Accuracy (%) | 91.1 |
| Tau2-Bench Telecom | 93.86 | Success Rate (%) | 91 |
| AA Omniscience - Software Engineering (SWE) - C | 60 | Accuracy (%) | 90.8 |
| Korean CSAT 2026 (Easy Mode) - English | 100 | Points (out of 100) | 90.5 |
Interactive version: theaggregate.ai/model?slug=kimi-k2-6-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.