Kimi K2.5 (Thinking) — benchmark results
Kimi K2.5 evaluated with thinking enabled. Provider: Moonshot. Released 2026-01-27. Access: Open.
Unified ELO 1788 ± 11, rank #153 of 1776 rated models, from 115 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UGI Leaderboard | 55.23 | UGI Score | 96.4 |
| AA TAU-2 Bench | 95.91 | Accuracy (%) | 96.2 |
| UGI - Natural Intelligence | 62.96 | NatInt Score | 95.8 |
| Vals AI CorpFin v2 | 68.26 | Accuracy (%) | 95.1 |
| UGI - Writing | 62.42 | Writing Score | 94.7 |
| Kagi LLM Benchmark | 78.5 | Accuracy (%) | 93.6 |
| BenchTable | 74.3 | Total Score (%) | 92.4 |
| LLMEval-Logic Formalization Fixed | 59.4 | Accuracy (%) | 92.3 |
| MathVision | 85 | Overall Accuracy (%) | 91.9 |
| AA Omniscience - Software Engineering (SWE) - Swift | 72 | Accuracy (%) | 91.7 |
| AA SciCode | 48.96 | Accuracy (%) | 91.1 |
| Vals AI AIME | 95.62 | Accuracy (%) | 91.1 |
Interactive version: theaggregate.ai/model?slug=kimi-k2-5-thinking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.