KAT-Coder-Pro V1
Intelligence
36
#117/532
Coding
18.3
#166/88
Math
94.7
#10/265
Speed
59 tok/s
TTFT: 1.25s
Pricing
$0.30 / $1.20
per 1M tokens (in/out)
KAT-Coder-Pro V1 is KwaiKAT’s model designed for coding and mathematical tasks. It processes at 58.767 tokens per second and is priced at $0.3 per million input tokens and $1.2 per million output tokens, targeting professional users.
When to Use KAT-Coder-Pro V1
✓ Best For
- Automating coding tasks.
- Performing complex mathematical calculations.
- Supporting developers in software engineering.
✗ Not Ideal For
- General-purpose language understanding tasks.
- High-speed real-time applications.
How KAT-Coder-Pro V1 Compares
Intelligence Index · Higher is better
DeepSeekChina MobileKwaiKATMiniMaxAnthropic
Benchmark Profile
Output Speed · tok/s
OpenAIMiniMaxKwaiKATAlibabaZ AI
Math Index
Z AIKimiKwaiKATOpenAIAmazon
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (13)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 36 | |
| Coding Index | 18.3 | |
| Math Index | 94.7 | |
| MMLU-Pro | 813% | |
| GPQA | 764% | |
| LiveCodeBench | 747% | |
| HLE | 334% | |
| SciCode | 36.6% | |
| IFBench | 68.4% | |
| LCR | 74% | |
| TerminalBench Hard | 9.1% | |
| Tau2 | 88.6% | |
| AIME 2025 | 94.7% |
Data: Artificial Analysis · Updated: March 27, 2026
Frequently Asked Questions (15)
When was KAT-Coder-Pro V1 released?
KAT-Coder-Pro V1 was released on November 11, 2025.
Who created KAT-Coder-Pro V1?
KAT-Coder-Pro V1 was created by KwaiKAT.
How intelligent is KAT-Coder-Pro V1?
KAT-Coder-Pro V1 scores 36 on the Artificial Analysis Intelligence Index, placing it well above average among other non-reasoning models in a similar price tier (median: 15).
How fast is KAT-Coder-Pro V1?
KAT-Coder-Pro V1 generates output at 27.5 tokens per second (based on KwaiKAT's API), which is at the lower end compared to other non-reasoning models in a similar price tier (median: 92.7 t/s).
What is the latency of KAT-Coder-Pro V1?
KAT-Coder-Pro V1 has a time to first token (TTFT) of 2.07s (based on KwaiKAT's API), which is somewhat higher than average compared to other non-reasoning models in a similar price tier (median: 1.35s).
How much does KAT-Coder-Pro V1 cost?
KAT-Coder-Pro V1 costs $0.30 per 1M input tokens (somewhat higher than average, median: $0.20) and $1.20 per 1M output tokens (somewhat higher than average, median: $0.70), based on KwaiKAT's API.
What is KAT-Coder-Pro V1 API pricing?
KAT-Coder-Pro V1 costs $0.30 per 1M input tokens and $1.20 per 1M output tokens (based on KwaiKAT's API). For a blended rate (3:1 input to output ratio), this is $0.53 per 1M tokens. Pricing may vary by provider.
How verbose is KAT-Coder-Pro V1?
When evaluated on the Intelligence Index, KAT-Coder-Pro V1 generated 4.5M output tokens, which is better than average compared to other non-reasoning models in a similar price tier (median: 5.5M).
Is KAT-Coder-Pro V1 a reasoning model?
No, KAT-Coder-Pro V1 is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.
What input modalities does KAT-Coder-Pro V1 support?
KAT-Coder-Pro V1 supports text input.
What output modalities does KAT-Coder-Pro V1 support?
KAT-Coder-Pro V1 supports text output.
Can KAT-Coder-Pro V1 process images?
No, KAT-Coder-Pro V1 does not support image input. It can only process text.
Is KAT-Coder-Pro V1 multimodal?
No, KAT-Coder-Pro V1 is not multimodal. It only supports text input.
What is the context window of KAT-Coder-Pro V1?
KAT-Coder-Pro V1 has a context window of 260k tokens. This determines how much text and conversation history the model can process in a single request.
Is KAT-Coder-Pro V1 open source?
No, KAT-Coder-Pro V1 is proprietary. The model weights are not publicly available.
