KAT Coder Pro V2: benchmark results

Kwaipilot's flagship agentic-coding KAT-Coder successor (MoE, 72B active). Provider: Kuaishou. Released 2026-03-27. Access: API.

Unified ELO 1633 ± 1, rank #172 of 1392 rated models, from 26 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Terminal-Bench Hard49.24Accuracy (%)94.5
AA TAU-2 Bench89.47Accuracy (%)83.9
AA IFBench66.67Accuracy (%)79.5
AA GPQA Diamond85.45Accuracy (%)78.8
Wolfram LLM Benchmarking Project53.8Correct Functionality (%)78
SnakeBench27TrueSkill Rating77
BenchmarkList ECI130.63Capability Index (ECI)76.1
AA Long Context Reasoning73Accuracy (%)73.2
Artificial Analysis Intelligence Index25.59Intelligence Index73.1
AA Humanity's Last Exam16.13Accuracy (%)67.7
SvelteBench91.1Average pass@1 (%)65
AA Omniscience-22.47Score62.2

Interactive version: theaggregate.ai/model?slug=kat-coder-pro-v2 · How It Works · Data refreshed daily, snapshot 2026-09-05.