KAT-Coder-Pro V1: benchmark results

Kwaipilot's (Kuaishou) first production agentic-coding KAT-Coder API model. Provider: Kuaishou. Released 2025-10-01. Access: API.

Unified ELO 1601 ± 1, rank #263 of 1392 rated models, from 42 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Wolfram LLM Benchmarking Project58.9Correct Functionality (%)86.7
AA Humanity's Last Exam33.64Accuracy (%)85.4
AA TAU-2 Bench88.6Accuracy (%)82.8
AA IFBench68.37Accuracy (%)82
AA Global-MMLU-Lite - Chinese88.42Accuracy (%)70.7
Artificial Analysis Intelligence Index21.69Intelligence Index67.7
BenchmarkList ECI119.61Capability Index (ECI)61.3
AA GPQA Diamond76.36Accuracy (%)59.7
AA Global-MMLU-Lite - Burmese77.33Accuracy (%)58.6
AA Global-MMLU-Lite - English90.75Accuracy (%)58.6
AA Global-MMLU-Lite - Korean84.42Accuracy (%)55.9
AA Global-MMLU-Lite - Indonesian86.33Accuracy (%)55.4

Interactive version: theaggregate.ai/model?slug=kat-coder-pro-v1 · How It Works · Data refreshed daily, snapshot 2026-09-05.