KwaiKAT
KAT-Coder-Pro V1
KAT-Coder-Pro V1 is a KwaiKAT language-model profile in the Easy Benchmarks snapshot. Use this page to compare its measured Artificial Analysis scores, output speed, time to first token, pricing, and relative ranking against other models in the local catalog.
KwaiKAT model releasesOperational Metrics
Output Speed117.1 tok/s
First Token1.28s
Blended Price$0.525/M
Model Metadata
Queryable facts extracted from the upstream model payload.
ReleaseNov 11, 2025
Context Windown/a
Modalitiesn/a
Strength: Math
Rank #10 across 269 models.
94.7
Strength: HLE
Rank #18 across 474 models.
33.4%
Strength: LCB
Rank #49 across 343 models.
74.7%
Strength Profile
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Benchmark Percentiles
Higher bars mean stronger relative placement.
All Benchmarks
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 36.0 | #100 |
| Artificial Analysis Coding Index | coding | 18.3 | #207 |
| Artificial Analysis Math Index | math | 94.7 | #10 |
| MMLU-Pro | reasoning | 81.3% | #90 |