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.
Aggregate Rating 1643 ± 41, rank #324 of 1656 rated models, from 73 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA AIME 2025 | 94.67 | Accuracy (%) | 96.1 |
| AA LiveCodeBench | 74.71 | Pass@1 (%) | 86.1 |
| AA TAU-2 Bench | 88.6 | Accuracy (%) | 82.8 |
| Tau2-Bench Telecom | 88.6 | Success Rate (%) | 82.6 |
| AA IFBench | 68.37 | Accuracy (%) | 82 |
| Wolfram LLM Benchmarking Project | 58.9 | Correct Functionality (%) | 81.7 |
| AA Humanity's Last Exam | 33.64 | Accuracy (%) | 79.9 |
| AA MMLU-Pro | 81.27 | Accuracy (%) | 73.7 |
| AA Global-MMLU-Lite - Chinese | 88.42 | Accuracy (%) | 70.5 |
| AA-Omniscience Hallucination Rate | 67.36 | Hallucination Rate (%) | 64.1 |
| Artificial Analysis Intelligence Index | 18.59 | Intelligence Index | 63.1 |
| AA-Omniscience Index - Software Engineering (SWE) - Julia | -28 | Omniscience Index | 60.9 |
Interactive version: theaggregate.ai/model?slug=kat-coder-pro-v1 · How It Works · Data refreshed daily, snapshot 2026-10-20.