K2-V2 (High) — benchmark results

Provider: LLM360. Released 2026-01-27. Access: Open.

Unified ELO 1593 ± 22, rank #510 of 1841 rated models, from 53 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA LiveCodeBench69.42Pass@1 (%)78.2
AA AIME 202578.33Accuracy (%)73.1
AA IFBench60.14Accuracy (%)71.4
AA MMLU-Pro78.59Accuracy (%)61
AA Humanity's Last Exam9.82Accuracy (%)60.8
AA Omniscience - Software Engineering (SWE) - Kotlin20Accuracy (%)54.3
AA-LCR33.3Score (self-reported)52.4
AA Omniscience - Health19.1Accuracy (%)49.7
AA GPQA Diamond68.08Accuracy (%)49.1
AA Omniscience - Science, Engineering & Mathematics25.8Accuracy (%)48.5
AA Omniscience - Software Engineering (SWE) - JavaScript29.09Accuracy (%)48.4
AA Omniscience - Software Engineering (SWE) - HTML30Accuracy (%)47.3

Interactive version: theaggregate.ai/model?slug=k2-v2-high · How It Works · Data refreshed daily, snapshot 2026-07-25.