Kimi K2.6 — benchmark results
Moonshot Kimi K2.6 open 1T-parameter MoE model. Provider: Moonshot. Released 2026-04-20. Access: Open.
Unified ELO 1749 ± 10, rank #199 of 1776 rated models, from 363 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LLM Stats (Claw-Eval) | 80.9 | Score (%) | 100 |
| LLM Stats (OJBench) | 60.6 | Score (%) | 100 |
| LLM Stats (WideSearch) | 80.8 | Score (%) | 100 |
| Slovenian OCR Benchmark | 98.65 | Score (%) | 100 |
| StructureClaw (Automatic Workflow) | 100 | Success Rate (%) | 100 |
| AGC-Bench - pron_vs_prompt | 1.19 | Dataset z-score | 98.8 |
| MathVision | 93.2 | Overall Accuracy (%) | 98.8 |
| JudgeBench Reasoning | 96.94 | Accuracy (%) | 98 |
| MATH-MC Level 1 | 99.3 | Accuracy (%) | 97.8 |
| OpenCompass Math - Competition | 72.1 | Score (%) | 97.7 |
| AGC-Bench - conceptual_design | 1.57 | Dataset z-score | 97.6 |
| AGC-Bench - hypogen | 1.95 | Dataset z-score | 97.6 |
Interactive version: theaggregate.ai/model?slug=kimi-k2-6 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.